--- Log opened Sun Nov 23 00:00:05 2008 20081123 00:01:04 mordante * r30990 /trunk/ (changelog src/font.cpp src/font.hpp src/game.cpp): 20081123 00:01:04 Fixed the font loading. 20081123 00:01:04 When the data directory is overridden the font path was initialized to 20081123 00:01:04 the initial value and never updated. 20081123 00:01:58 night 20081123 00:02:33 -!- mordante [n=chatzill@roadie.xs4all.nl] has quit ["ChatZilla 0.9.84 [Firefox 3.0.3/2008092417]"] 20081123 00:11:47 -!- loonycyborg [n=sergey@79.139.247.220] has quit ["Zzzzzzzzzzzzzzzzzzzzzzz"] 20081123 00:12:25 -!- loonybot [n=loonybot@79.139.247.220] has quit [Remote closed the connection] 20081123 00:23:56 zookeeper * r30991 /trunk/data/core/ (3 files in 2 dirs): Added two silver mage teleport frames that were mysteriously missing from the proper commit. 20081123 00:25:00 [base_unit] probably needs an attribute inherit_animations=true/false 20081123 00:25:27 zookeeper: would that be useful? 20081123 00:25:44 -!- VladimirSlavik [n=chatzill@136.219.broadband6.iol.cz] has left #wesnoth-dev [] 20081123 00:27:30 i'm not sure, really. merging of tags like in gender variations is already something i'd rather not see. 20081123 00:29:12 that's exactly what's happening with the arch mage's idle animation 20081123 00:29:31 yep 20081123 00:30:51 I think merging of tags is not the problem, but rather merging of animation tags in variants which wish to use different animations 20081123 00:31:18 (or no animations) 20081123 00:32:38 I could make the attribute inherit_animations=true/false available for [female] also 20081123 00:36:43 I can't get UMC translations to work =/ 20081123 00:36:55 TDH has a whole bunch of .mo files in its translations dir 20081123 00:37:09 no location found for 'data/campaigns/The_Dark_Hordes/translations', not adding textdomain 20081123 00:37:14 but that's what wesnoth says 20081123 00:37:25 I've moved them around, but that didn't help 20081123 00:53:02 -!- Elvish_Pillage2 [n=eli@24-177-39-196.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev 20081123 01:05:37 ai0867 * r30992 /trunk/data/core/images/ (34 files in 3 dirs): Ran set-properties on /data/core/images. 20081123 01:06:34 -!- Espreon [n=espreon@c-76-117-239-33.hsd1.pa.comcast.net] has joined #wesnoth-dev 20081123 01:07:05 ai0867 * r30993 /trunk/data/core/music/ (heroes_rite.ogg love_theme.ogg the_deep_path.ogg): Ran set-properties on /data/core/music. 20081123 01:09:34 -!- Elvish_Pillager [n=eli@24-177-39-196.dhcp.oxfr.ma.charter.com] has quit [Read error: 110 (Connection timed out)] 20081123 01:18:29 i am off to bed, n8 20081123 01:19:22 -!- Blueblaze [n=Blueblaz@c-98-196-208-51.hsd1.tx.comcast.net] has joined #wesnoth-dev 20081123 01:35:51 -!- [Relic] [n=[Relic]@adsl-76-229-220-201.dsl.milwwi.sbcglobal.net] has quit ["Leaving"] 20081123 01:40:23 esr: 01:39 < AI0867> wmllint just complained about a newline in a string?? 20081123 01:40:23 01:39 < AI0867> TDH/scenarios/09:235 20081123 01:40:23 01:39 < AI0867> and yes, there is a newline there, but why is that a problem? 20081123 01:45:18 -!- fendrin_ [n=fabi@g228067232.adsl.alicedsl.de] has joined #wesnoth-dev 20081123 01:46:43 Sapient, yeah, sounds like an ok idea 20081123 01:47:01 AI0867: really? 20081123 01:47:12 this sounds like a blocker. 20081123 01:47:17 wow, delayed response there zookeeper 20081123 01:47:35 (regarding translations not working /being detected on UMC) 20081123 01:48:53 -!- thespaceinvader [n=chatzill@91.108.187.175] has quit ["ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]"] 20081123 01:50:16 * Shadow_Master grins 20081123 01:51:25 Sapient, sometimes we are slow 20081123 01:51:35 -!- zookeeper [n=l@dsl-tkubrasgw1-fe11fa00-51.dhcp.inet.fi] has quit [] 20081123 01:51:56 is it just me or the forums are being infested with image macros... 20081123 01:52:24 Hmmmmm? 20081123 01:52:31 if you ask for trolls, you get them 20081123 01:52:58 * Espreon 's little brothers are Troll Whelps. 20081123 01:54:55 be a bigger troll and annoy them until they go away ;) 20081123 01:55:31 * Shadow_Master scratches head 20081123 01:57:21 * Espreon wishes that he could use his club attack w/o getting in trouble if he would pretend to be a Troll. 20081123 01:57:41 I got defeated at the first attempt of playing UtBS scenario 1 on EASY 20081123 01:57:56 by a DA 20081123 01:58:14 -!- [Relic] [n=[Relic]@adsl-76-229-220-201.dsl.milwwi.sbcglobal.net] has joined #wesnoth-dev 20081123 01:58:40 this is so frustrating that it reminds me why I don't play Wesnoth nowadays :( 20081123 01:59:07 -!- fendrin [n=fabi@g226199156.adsl.alicedsl.de] has quit [Read error: 110 (Connection timed out)] 20081123 01:59:16 <[Relic]> Hello :) 20081123 01:59:26 whoah, the new message dialog has positively tiny portraits these days 20081123 01:59:30 I almost missed the lich 20081123 02:00:06 -!- boucman [n=rosen@159.83.102-84.rev.gaoland.net] has quit [Remote closed the connection] 20081123 02:05:25 gah! I can't remember by SF password 20081123 02:05:38 there must be some way they can send/reset it for you 20081123 02:06:31 crimson_penguin: Think deeply. 20081123 02:07:07 Gah! And they have a VERY old email 20081123 02:09:34 And I don't have a security question... 20081123 02:09:55 :O 20081123 02:14:54 AI0867: I think it warns about newlines in case you tried to manually word-wrap your paragraphs, instead of allowing the engine to do that. 20081123 02:16:06 I think it needs to be a bit smarter about it if it is going to try to detect such a thing 20081123 02:16:19 Shadow_Master: luckily, they have a recovery thing, just for my situation... but it's not automatic, so I'll have to wait, and hope they believe it's really me 20081123 02:20:13 I got it! 20081123 02:20:47 Yay! 20081123 02:20:51 * crimson_penguin used leet javascript skills 20081123 02:20:52 ai0867 * r30994 /trunk/src/ (unit.cpp unit_types.cpp): Remove obsolete compatibility code. 20081123 02:21:08 I manually wordwrapped it because I wanted it to be on different lines 20081123 02:21:23 That was really stupid - it was actually my newest password, I just didn't even bother to try it because I thought if it had an email that old it couldn't possibly have that password 20081123 02:21:36 AI0867: can you paste the text in question? 20081123 02:21:56 You can suppress warnings about newlines in messages (and attempts to 20081123 02:21:56 # repair them) with "wmllint: display on", and re-enable them with 20081123 02:21:56 # "wmllint: display off" 20081123 02:22:59 * crimson_penguin went to the login page and "went to": javascript:alert(document.getElementsByTagName("input")[6].value); return false; 20081123 02:23:02 message= _ "Block the entrances. We will be safe here for a while. 20081123 02:23:05 Now I have to study the book." 20081123 02:24:03 is there an empty line after "while" ? 20081123 02:24:14 no 20081123 02:24:24 maybe there should be 20081123 02:24:27 maybe I should put one there 20081123 02:27:28 the warning could certainly stand to be clearer what it is warning about 20081123 02:27:51 Mac binary released :) 20081123 02:28:05 Yay. 20081123 02:28:27 Added to the SF release, that is 20081123 02:28:34 And I better get back to the party next door 20081123 02:29:04 Ivanovic: I added my dmg to the 1.5.6 release - hopefully I did everything right, and hopefully it works :) 20081123 02:29:24 Sapient: adding that newline didn't remove the warning 20081123 02:29:41 AI0867: yeah, like I said, the check needs to be smarter 20081123 02:30:39 if you added the newline and the warning still fired, then apparently it doesn't recognize paragraphs 20081123 02:30:51 better turn off my computer now 20081123 02:30:53 hejdå 20081123 02:30:56 -!- crimson_penguin [n=ben@64.201.60.217] has quit ["Leaving"] 20081123 02:33:39 AI0867: I will commit a fix to it in a sec... 20081123 02:34:40 k 20081123 02:35:06 as for the UMC translations thing, it looks for the path specified in all binary_paths 20081123 02:35:51 or that's what it seems to do 20081123 02:35:52 testing now 20081123 02:37:29 okay, setting the textdomain path to 'translations' got rid of the error 20081123 02:37:40 not that any translations actually showed up, but still... 20081123 02:38:45 but considering the textdomain path added to utbs when moved to branches (data/campaigns/Under_The_Burning_Suns/translations), this isn't how it should work 20081123 02:40:21 alink changed it in 27176 20081123 02:41:08 that revision was when path= was added... 20081123 02:42:55 AI0867: update svn and try it again please 20081123 02:45:08 shikadibot: log head 20081123 02:45:10 Revision 30995 (sapient 2008-11-23 01:43:15 +0000 (Sun, 23 Nov 2008)): 20081123 02:45:13 be a little smarter about newlines in messages 20081123 02:45:15 Web interface URL: http://svn.gna.org/viewcvs/wesnoth?view=rev&rev=30995 20081123 02:45:53 -!- Elvish_Pillage2 [n=eli@24-177-39-196.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!"] 20081123 02:46:19 hmm... is that bot open source? :) 20081123 02:46:22 it is 20081123 02:46:29 yey. could I get the source? 20081123 02:46:46 http://wesnoth-umc-dev.svn.sourceforge.net/svnroot/wesnoth-umc-dev/branches/irc-bots/shikadibot 20081123 02:46:51 thankie 20081123 02:46:59 Have fun. 20081123 02:47:04 Sapient: that fixed it 20081123 02:47:26 good 20081123 03:03:32 -!- crimson_penguin [n=ben@64.201.60.217] has joined #wesnoth-dev 20081123 03:06:55 sapient * r30995 /trunk/data/tools/wmllint: be a little smarter about newlines in messages 20081123 03:07:23 apparently the party couldn't go on without me; it was over when I went back :P 20081123 03:13:52 AI0867: is TDH all shiny and new ready to be tested? 20081123 03:16:55 -!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev 20081123 03:22:34 hmm, does it make sense to have password protected games colored in red? 20081123 03:22:46 i just noticed that's how it works in 1.5. 20081123 03:25:24 seems confusing to me since red is for unobservable games and you don't even need the password to observe password protected games. 20081123 03:26:40 Soliton: it is 20081123 03:26:56 also, about TSG: http://www.wesnoth.org/forum/viewtopic.php?f=8&t=2998&p=322858#p322858 20081123 03:28:06 well, you have to see the bandit leader maybe? 20081123 03:28:18 probably 20081123 03:28:22 afair anyway. 20081123 03:28:29 doesn't sound like he did that. 20081123 03:28:33 but I believe Ethiliel insists on killing the bandits too? 20081123 03:29:52 i think so. 20081123 03:35:30 the fake lich's die event doesn't check for that though 20081123 03:39:01 wow, there hasn't been a Mac release since 1.5.3 20081123 03:39:06 ai0867 * r30996 /trunk/data/campaigns/The_South_Guard/scenarios/09a_Vengeance.cfg: Do not disable the winning event if it's triggered too early. 20081123 03:39:27 shikadibot: log 30996 20081123 03:39:28 Revision 30996 (ai0867 2008-11-23 02:39:36 +0000 (Sun, 23 Nov 2008)): 20081123 03:39:28 Do not disable the winning event if it's triggered too early. 20081123 03:39:28 Web interface URL: http://svn.gna.org/viewcvs/wesnoth?view=rev&rev=30996 20081123 03:41:44 how come shikadibot isn't listed in the bots section of the credits? ;) 20081123 03:42:00 it's not an 'official' bot =P 20081123 03:43:42 ah :P 20081123 03:43:58 also, I should maybe be listed as a packager next time? :) 20081123 03:45:39 go add yourself to data/core/about.cfg then 20081123 03:47:42 well... my svn checkout is anonymous, even though I do have commit access... 20081123 03:47:47 -!- Espreon [n=espreon@c-76-117-239-33.hsd1.pa.comcast.net] has quit [Read error: 110 (Connection timed out)] 20081123 03:48:15 k, what should I call you in your entry? 20081123 03:48:39 Ben Anderman (crimson_penguin)? 20081123 03:48:46 that'd do it :) 20081123 03:49:47 any email I should add? 20081123 03:50:02 there's emails in there? 20081123 03:50:12 optionally 20081123 03:50:19 usually spam-armored 20081123 03:50:28 ben at happyspork.com 20081123 03:50:41 I didn't see any emails in the in-game credits though 20081123 03:50:56 they're not in there 20081123 03:51:01 oh, ok 20081123 03:51:01 they end up on the wiki's about page though 20081123 03:51:12 ah, ok, makes sense 20081123 03:51:27 I'm not too weird about spam anyway; I have good filtering (black-list based) 20081123 03:51:47 not the coolhemp one? 20081123 03:52:22 nope, that one still works for now, but should be phased out 20081123 03:52:44 the future of my coolhemp address (and of coolhemp itself) is uncertain ;) 20081123 03:52:52 adding as ben_AT_happyspork.com (spaces break the wikifier) 20081123 03:53:14 ok, great 20081123 03:53:26 whereas happyspork.com I control, and will keep it going indefinitely :) 20081123 03:53:31 ai0867 * r30997 /trunk/data/core/about.cfg: Add crimson_penguin to about.cfg as mac packager. 20081123 03:53:38 thanks! 20081123 03:59:04 wesnoth: src/sdl_utils.cpp:1343: void blit_surface(const surface&, const SDL_Rect*, surface&, const SDL_Rect*): Assertion `dst_rect.x >= 0' failed. <-- not-reproducible crash 20081123 04:05:51 * crimson_penguin wonders why dst_rect.x should be >= 0 20081123 04:06:14 I thought SDL supported a negative value there; it would certainly make things simpler in some cases 20081123 04:11:50 AI0867: "File README is empty." 20081123 04:21:21 I haven't removed that yet? 20081123 04:21:52 I think I was planning to "maybe add something someday" 20081123 04:23:14 -!- PK2 [n=pk@r74-192-44-206.vctrcmta01.vctatx.tl.dh.suddenlink.net] has quit ["Java user signed off"] 20081123 04:23:38 -!- PK [n=pk@r74-192-44-206.vctrcmta01.vctatx.tl.dh.suddenlink.net] has joined #wesnoth-dev 20081123 04:29:15 https://gna.org/bugs/index.php?12317 <-- remind me, did we want to fix this or not? 20081123 04:38:15 would be nice. 20081123 04:39:10 notepad on windows seems to use such a format by default. 20081123 04:43:38 most windows apps do 20081123 04:44:15 which of course, breaks utf-8's nice feature of being identical to ASCII if no non-ASCII characters are used 20081123 04:49:59 -!- Psyche^ [n=Psyche@e177234042.adsl.alicedsl.de] has joined #wesnoth-dev 20081123 04:52:36 -!- Ivanovic_ [n=ivanovic@dtmd-4db2a2ba.pool.einsundeins.de] has joined #wesnoth-dev 20081123 04:53:48 -!- Patterner [n=Psyche@e177113157.adsl.alicedsl.de] has quit [Read error: 110 (Connection timed out)] 20081123 04:53:49 -!- Psyche^ is now known as Patterner 20081123 04:56:00 hi 20081123 04:56:19 did Ivanovic do a release? 20081123 04:56:59 is the repository free for commits again? 20081123 04:57:12 yes. 20081123 05:05:10 read the topic ;) 20081123 05:09:55 -!- Ivanovic [n=ivanovic@dtmd-4db2a30f.pool.einsundeins.de] has quit [Read error: 110 (Connection timed out)] 20081123 05:10:34 -!- Ivanovic_ is now known as Ivanovic 20081123 05:17:19 what can I do to prevent a unit from dying in a die event? 20081123 05:17:34 give it life 20081123 05:18:38 -!- Appleman1234 [n=Appleman@CPE-124-177-91-162.qld.bigpond.net.au] has joined #wesnoth-dev 20081123 05:18:46 heal it 20081123 05:19:01 thanks 20081123 05:20:51 PK: in version 1.5, if you unstore a unit with negative HP, it will get 1 hp (I think) 20081123 05:21:32 although I don't think this is a great idea, it's what the wiki states 20081123 05:22:13 -!- noy [n=Noy@d75-157-52-251.bchsia.telus.net] has joined #wesnoth-dev 20081123 05:27:39 so, what's the approximate date for 1.4.7? 20081123 05:31:39 -!- Espreon [n=espreon@c-76-117-239-33.hsd1.pa.comcast.net] has joined #wesnoth-dev 20081123 05:41:35 -!- Sapient [n=sapien-x@c-71-59-39-246.hsd1.ga.comcast.net] has quit [] 20081123 05:46:37 AI0867: You did wrote the advance_unit macro? 20081123 05:47:23 fendrin_: May we discuss UtBS? 20081123 05:47:31 fendrin_: yes 20081123 05:48:08 Espreon: wait, I am working Delfadors Memoirs right now 20081123 05:48:12 ugh, I was trying to fix bug #12596 20081123 05:48:19 removing the network player is easy 20081123 05:48:31 AI0867: can you have a look why it's not updating the graphic in a scenario of DM? 20081123 05:48:38 unfortunately the rest of the code seems to be using hardcoded offsets 20081123 05:48:38 fendrin_: But are you still the official maintainer of UtBS or did esr/zookeeper take the role? 20081123 05:49:11 I'll see if I can fix those in the morning 20081123 05:49:23 fendrin_: k 20081123 05:49:56 Espreon: I'm not sure about this. Fact is that I have no idea what esr/zookeeper did to utbs so I am not competent at the moment. 20081123 05:50:22 AI0867: I do commit my changes 20081123 05:51:20 fendrin_: which scenario 20081123 05:51:30 AI0867: 10, commited 20081123 05:51:51 fendrin_: kill the enemy leader with spawned debug units 20081123 05:53:26 k 20081123 05:54:17 fendrin_: I think they were just giving UtBS some life support. 20081123 05:54:30 I know this is the case with esr.. but IDK about zookeeper. 20081123 05:54:50 Espreon: what do you want to discuss? 20081123 05:55:17 Well... I wanted to discuss the minions of the Iron Council. 20081123 05:55:36 whoah 20081123 05:55:39 that's strange 20081123 05:55:46 indeed 20081123 05:56:08 the teleport macro seems to not work in this case as well 20081123 05:56:32 he's in his proper location now 20081123 05:56:41 but yes, teleport seems to behave strangely 20081123 05:56:42 I think their sides should use the undead flags since they are affiliated with the Iron Council (who uses the undead flag). 20081123 05:58:33 I'm off 20081123 05:58:40 Espreon: yeah, I like the idea of custom flags 20081123 05:59:22 AI0867: good bye 20081123 06:00:44 Also, I am gonna do this with Yechnagoth's minions on Zocthanol Island... all right? 20081123 06:03:36 yes 20081123 06:05:53 I am not sure whether or not to assign it to the traitor elves... as they would be asscociated with Chaos in a way as well... 20081123 06:05:57 Probably not... 20081123 06:06:31 no, they don't know that their god is a false one 20081123 06:06:50 they believe to act good 20081123 06:07:32 Makes sense. 20081123 06:10:45 espreon * r30998 /trunk/data/campaigns/Under_the_Burning_Suns/scenarios/ (3 files): * Assigned the undead flag to sides associated with Yechnagoth and the Iron Council, except for traitor elves. 20081123 06:10:54 I would like to have the campain selection dialog sort them by timeline. 20081123 06:12:34 But hasn't the soft string freeze already occurred? 20081123 06:12:53 Well... we could just throw something in branches or on umc-dev or something. 20081123 06:15:39 * PK has a question 20081123 06:16:05 When coding an [attack] in the [unit] is it possible to define a new "range=" 20081123 06:16:13 other than melee/ranged 20081123 06:17:03 sure, there is nothing hard coded about those. 20081123 06:18:11 Espreon: for what change would you like to brake the string freeze? 20081123 06:19:00 -!- crimson_penguin [n=ben@64.201.60.217] has quit ["Leaving"] 20081123 06:20:56 Soliton: also, is it possible to define a range in special ways? like say everyone fighting in range "longrange" has -10% cth 20081123 06:21:13 or do I have to code a special to give all attacks in "longrange" 20081123 06:21:28 fendrin_: I want Kaleh to mention the fact that the people "Eloh" wants them to go with has those flags. 20081123 06:21:57 that'd have to be a weapon special, i guess. 20081123 06:26:35 I am not sure if the people ingame do see this flags. 20081123 06:27:21 Well... the people in IFtU ostensibly do. 20081123 06:29:42 Espreon: you can insert this mention but comment it out. After the 1.6 release you comment it in again. 20081123 06:30:02 Don't you mean uncomment it? 20081123 06:30:20 yeah 20081123 06:39:36 -!- Appleman1234 [n=Appleman@CPE-124-177-91-162.qld.bigpond.net.au] has quit [Read error: 110 (Connection timed out)] 20081123 06:41:28 fendrin_: Maybe we should throw something in branches or in umc-dev so we can work on the story and dialogue. 20081123 06:43:42 Espreon: good idea 20081123 06:46:07 I guess we could update the thing in /branches/utbs and work on the story/dialogue in there. 20081123 06:49:49 and merge it to trunk after the release of 1.6 20081123 06:50:09 More like "merge." 20081123 06:50:22 True merging is scary as I heard. 20081123 06:50:27 *,as 20081123 06:50:33 *, as 20081123 06:53:14 That depends. If we are only doing bug fixes in the trunk version the merging task shouldn't be too complicated. 20081123 06:53:35 I meant true merging in SVN. 20081123 06:53:54 And of course we would only fix bugs in trunk now. 20081123 06:58:12 All right... this is gonna be a bit tricky. 20081123 06:58:37 Meh... I do not feel like playing with SVN right now... 20081123 07:01:12 Is anyone here an SVN expert? 20081123 07:09:54 "src/actions.cpp:2555: void apply_shroud_changes(undo_list&,...): Assertion 'unit_itor != units.end()' failed." UtBS:OotFP 20081123 07:10:18 * Patterner needs just "svn co" and "up" :) 20081123 07:11:23 Hopefully. 20081123 07:15:25 -!- [Relic] [n=[Relic]@adsl-76-229-220-201.dsl.milwwi.sbcglobal.net] has quit ["Leaving"] 20081123 08:28:55 -!- PK [n=pk@r74-192-44-206.vctrcmta01.vctatx.tl.dh.suddenlink.net] has quit ["Java user signed off"] 20081123 08:56:39 -!- Psyche^ [n=Psyche@e177227009.adsl.alicedsl.de] has joined #wesnoth-dev 20081123 09:08:52 -!- Patterner [n=Psyche@e177234042.adsl.alicedsl.de] has quit [Read error: 110 (Connection timed out)] 20081123 09:08:53 -!- Psyche^ is now known as Patterner 20081123 09:28:39 -!- boucman [n=rosen@159.83.102-84.rev.gaoland.net] has joined #wesnoth-dev 20081123 09:35:11 -!- Sirp [n=user@c-76-102-104-121.hsd1.ca.comcast.net] has quit ["leaving"] 20081123 09:43:55 -!- Blueblaze [n=Blueblaz@c-98-196-208-51.hsd1.tx.comcast.net] has quit [] 20081123 10:06:47 -!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has joined #wesnoth-dev 20081123 10:18:12 -!- Patterner [n=Psyche@e177227009.adsl.alicedsl.de] has quit [Read error: 110 (Connection timed out)] 20081123 10:18:54 -!- Patterner [n=Psyche@e177114027.adsl.alicedsl.de] has joined #wesnoth-dev 20081123 10:23:21 -!- zookeeper [n=l@dsl-tkubrasgw1-fe11fa00-51.dhcp.inet.fi] has joined #wesnoth-dev 20081123 10:24:43 zookeeper: Hello. 20081123 10:24:52 Sadly nobody responded to the quest... yet. 20081123 10:24:59 *request 20081123 10:25:22 * Espreon kills the yet. 20081123 10:26:52 -!- mordante [n=chatzill@roadie.xs4all.nl] has joined #wesnoth-dev 20081123 10:27:10 morning 20081123 10:27:35 mordante: Hello. 20081123 10:27:50 hi Espreon 20081123 10:27:55 hello 20081123 10:28:03 hi zookeeper 20081123 10:28:17 morning all 20081123 10:28:25 hi boucman 20081123 10:28:33 Hello boucman. 20081123 10:28:39 and no, i'm not going to start maintaining UtBS or anything...i was just helping get rid of some bugs. i don't think esr is going to start maintaining it either. 20081123 10:28:59 Ah... so both of you were just giving it some life support. 20081123 10:29:26 * Espreon hopes that someone will respond to the request soon. 20081123 10:29:45 yes 20081123 10:29:56 what request ? 20081123 10:31:02 The request to recolor the Elvish Hunters from IFtU to be used as its Quenoth counterpart. In addition we need its L2 and L3 animated. 20081123 10:31:08 *Hunter 20081123 10:31:23 * boucman is really not an artist 20081123 10:32:23 If nobody responds... I will try recoloring the L1... but animating the L2 and the L3... is gonna be someone else's job. 20081123 10:34:23 Maybe it is because my username is not in bold and yellowish green. 20081123 10:35:31 Only 2207 more packages to install. 20081123 10:35:44 *install/update 20081123 10:36:43 why do you need that colour? maintainers tend to get bold red ;) 20081123 10:37:26 It is Jetryl's color. 20081123 10:37:51 yes and once you have that colour people ask, why can't you do it yourself ;) 20081123 10:38:01 btw do you read the dev-ml? 20081123 10:38:12 * Espreon is the Skittle Lord... thus he commands. 20081123 10:38:27 Plus I would be the Art DIRECTOR. 20081123 10:38:32 I would be directing. 20081123 10:39:03 moin 20081123 10:39:17 hi Ivanovic 20081123 10:39:26 Ivanovic: Hello. 20081123 10:40:29 mordante: In addition, I have seen Jetryl make requests. 20081123 10:40:45 I know ;) 20081123 10:41:04 but you might ask elezar 20081123 10:41:29 * Espreon has the crazy idea that his username's color should be the same as Jetryl's until the request has been fulfilled. 20081123 10:41:30 Ivanovic any reason Espreon didn't get his red shirt on the forum? 20081123 10:41:43 Hmmm? 20081123 10:42:01 you're the maintainer of UtBS or not? 20081123 10:42:05 Co. 20081123 10:42:22 Espreon, the co-maintainer... 20081123 10:43:03 mordante: he has not told me his form account name... 20081123 10:43:53 ah good reason 20081123 10:44:13 Ivanovic: My forum username... is.... -drum roll-... Espreon. 20081123 10:44:33 the nick of fendrin_ was 'fabi', right? 20081123 10:44:37 (since he is listed) 20081123 10:44:40 Yup. 20081123 10:45:39 now you are in the group of all evil 20081123 10:45:50 Yay... hail the Dark Lady! 20081123 10:47:14 now you can see whether your new status helps you to get the images recoloured ;) 20081123 10:47:34 Hopefully it will. 20081123 10:49:05 * Espreon always wanted his user title to be: Lord of Skittles 20081123 10:49:25 NO WAY!!! 20081123 10:49:33 I knew this would happen. 20081123 10:49:42 It was worth a shot. 20081123 10:50:00 Now to continue watching anime as my packages install/update. 20081123 11:06:39 -!- Netsplit niven.freenode.net <-> irc.freenode.net quits: yann 20081123 11:08:32 -!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev 20081123 11:09:39 Say No to Eloh! 20081123 11:10:35 Don't you mean "Eloh"? 20081123 11:10:45 Eloh is different than "Eloh". 20081123 11:10:57 Say No to "Eloh" too! 20081123 11:11:05 Eloh is Elynia, on the other hand "Eloh" is Yechnagoth. 20081123 11:11:28 Don't confuse me with facts. 20081123 11:11:42 * Espreon likes to confuse people with facts. 20081123 11:11:57 Say NO to Espreon :P 20081123 11:12:27 * Espreon cries. 20081123 11:13:15 * Patterner gives Espreon a durian cookie 20081123 11:13:26 * Espreon refuses. 20081123 11:19:51 -!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has joined #wesnoth-dev 20081123 11:26:42 -!- Shadow_Master [n=shadowm@unaffiliated/shadowmaster/x-275924] has quit [Read error: 60 (Operation timed out)] 20081123 11:27:39 Ivanovic: the nick of fabi is fendrin 20081123 11:27:56 -!- AI0867 [n=ai@van-gessel.demon.nl] has quit [Read error: 60 (Operation timed out)] 20081123 11:29:43 fendrin_: lets say it a little different: 20081123 11:30:11 the forum account 'fabi' belongs to the same dude that has commit access and tends to call himself in here 'fendrin' 20081123 11:30:13 correct? 20081123 11:30:23 indeed 20081123 11:31:47 that is already mentioned in the wiki 20081123 11:33:19 -!- Elvish_Pillager [n=eli@24-177-39-196.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev 20081123 11:37:26 i don't trust the wiki, everyone can do changes in there... 20081123 11:41:21 that is true. I assumed that you wanted to have the information to write it into the wiki. 20081123 11:41:34 -!- shikadibot [n=ai@van-gessel.demon.nl] has quit [Read error: 110 (Connection timed out)] 20081123 11:41:41 -!- stikonas [n=and@ctv-213-164-107-170.vinita.lt] has joined #wesnoth-dev 20081123 11:44:15 ivanovic * r30999 /trunk/data/campaigns/Legend_of_Wesmere/ (_main.cfg about.cfg): 20081123 11:44:15 putting about.cfg into an external file does *not* work with the data extraction scripts 20081123 11:44:15 moving the content into the _main-cfg file and removing the "old" files 20081123 11:47:54 okay, finally the credits on the website are up to date again 20081123 11:56:07 -!- miq [n=miq@krlh-5d83b435.pool.einsundeins.de] has joined #wesnoth-dev 20081123 11:57:05 -!- noy [n=Noy@d75-157-52-251.bchsia.telus.net] has quit [] 20081123 11:57:42 Ivanovic: there is something wrong with the credits on the website. Look at: http://www.wesnoth.org/wiki/Credits#Macedonian_Translation 20081123 12:00:34 stikonas: no, it is not wrong 20081123 12:00:35 in fact, there is no translator for Macedonian in data/core/about.cfg 20081123 12:00:41 there is just noone liste for this translation 20081123 12:01:11 with other words: the files are there but that's it 20081123 12:01:20 noone really has worked on it so far (at least for mainline) 20081123 12:01:30 i think mythological asked for this translation to be added 20081123 12:01:59 Hmm I get a compilation error for 1.5.6... "po/wesnoth-low/racv.po not found" 20081123 12:02:35 you are right, the file is not there... 20081123 12:03:22 I just did scons --jobs 4 prefix=/foo 20081123 12:03:54 uhm, no, the file *is* there on my box 20081123 12:04:24 and it is under version control, too 20081123 12:04:35 I got the tarball from sf.net 20081123 12:04:56 3eefd2c0e0129677bf0da66044fe5d1d wesnoth-1.5.6.tar.bz2 20081123 12:05:00 md5sum 20081123 12:07:10 found the problem 20081123 12:07:18 but no, i will not reupload the tarball 20081123 12:07:23 Ah, I just remembered something. 20081123 12:07:26 ivanovic * r31000 /trunk/po/wesnoth-low/LINGUAS: racv missing in linguas file resulting in file not shipping in tarball... 20081123 12:07:31 (with autotools you got no problem building...) 20081123 12:07:52 I thought autotools wasn't "supported" in 1.5 anymore? 20081123 12:08:28 busfahrer: it is still working anyway 20081123 12:08:35 it is still supported 20081123 12:08:37 and yes, cmake aborts with the same error since the file is not there 20081123 12:08:48 Let me try, with 1.5.5 I got an error when using autotools 20081123 12:08:52 not supported as in "when it finally breaks we will not fix it) 20081123 12:08:58 hehe 20081123 12:09:22 and I'm really tempted to remove the deprecated message as well since nobody is willing to start a real evaluation 20081123 12:11:53 Ivanovic: I want to use /branches/utbs/ as a place to develop the story, improve dialogue, etc. for UtBS... so.. would it be acceptable to empty its contents and refill the directory with the contents of /trunk/data/campaigns/Under_the_Burning_Suns in two commits? 20081123 12:12:16 * Espreon is not an SVN guru... thus he does not know all the tricks of SVN. 20081123 12:12:34 Espreon: sure 20081123 12:12:46 All right... here comes the fun part... 20081123 12:14:29 espreon * r31001 /branches/utbs/ (7 files in 7 dirs): * Emptied /branches/utbs/ in order to make updating much easier. 20081123 12:17:23 ….. 20081123 12:17:30 espreon * r31002 /branches/utbs/ (TODO _main.cfg): * Updated /branches/utbs to the latest version in trunk. 20081123 12:17:54 Looks like I am going to have to use svn add on individual files. 20081123 12:18:13 Espreon: svn cp should do 20081123 12:18:31 fendrin_: Thanks. 20081123 12:18:47 Espreon: uhm, try this: svn rm svn+ssh://NICKNAME@svn.gna.org/svn/wesnoth/branches/utbs 20081123 12:19:07 Ivanovic: Thank you very much. 20081123 12:19:15 svn cp svn+ssh://NICKNAME@svn.gna.org/svn/wesnoth/trunk/data/campaigns/Under_the_Burning_Suns svn+ssh://NICKNAME@svn.gna.org/svn/wesnoth/branches/utbs 20081123 12:19:21 and just enter your nickname 20081123 12:19:54 fendrin * r31003 /branches/utbs/: cleaned the branches/utbs tree 20081123 12:20:24 I will do the rest. 20081123 12:20:28 okay 20081123 12:22:03 espreon * r31004 /branches/utbs/: * Updated /branches/utbs/ to the latest version in trunk. 20081123 12:23:45 mordante: the message looks totally fine to me 20081123 12:23:48 fendrin_: All right, now that we have the sandbox set up... we may start improving the story and whatnot. 20081123 12:23:53 regarding autotools and cmake/scons 20081123 12:24:48 well who's evaluating scons/cmake ? 20081123 12:25:45 my last tests with scons where no joy, my biggest problem is to change my workflow and I'm not really in the mood to invest time in that 20081123 12:26:06 make file.o and make are just so simple 20081123 12:26:36 and still didn't manage to get ccache properly working 20081123 12:26:42 what about cmake? 20081123 12:27:22 also ran into problems with ccache but at least it still uses make 20081123 12:27:45 but like I said I don't want to spend too much time on it, rather spend time on coding 20081123 12:27:52 ;) 20081123 12:28:26 Espreon: my first task will be to remove the hardcoded flood in scenario 4 and replace it with a recursive function. 20081123 12:29:21 -!- stikonas [n=and@ctv-213-164-107-170.vinita.lt] has quit [Read error: 60 (Operation timed out)] 20081123 12:29:24 also I've been helping to maintain autotools so I see no reason to remove it as long as we have people willing to maintain it 20081123 12:29:58 fendrin_: Be sure to do it in trunk, as the purpose of the sandbox was to get around the string freeze. 20081123 12:30:48 Espreon: right. that can go to trunk 20081123 12:33:38 esr * r31005 /trunk/data/campaigns/ (2 files in 2 dirs): 20081123 12:33:38 Move a unit declaration that was dodgy because (a) a macro it used was 20081123 12:33:38 questionable within [side], and (b) wmllint got confused. 20081123 12:38:08 ivanovic * r31006 /trunk/ (README config.h.dummy configure.ac): 20081123 12:38:08 correct some values 20081123 12:38:08 rewrite the "evaluating build systems"-part to refelect current situation better 20081123 12:38:08 insert correct places for reporting bugs 20081123 12:38:30 mordante: here ccache seems to work "automatically" with cmake 20081123 12:38:50 that is: sure, with the first build the cache is regenerated since a different compiler command will be used 20081123 12:39:02 mordante: is this description better? 20081123 12:41:14 -!- loonycyborgw32 [i=kvirc@79.139.247.220] has joined #wesnoth-dev 20081123 12:44:28 Ivanovic yes the text looks better, maybe I used an older version of cmake when I had ccache problems 20081123 12:45:06 ivanovic * r31007 /trunk/ (4 files in 3 dirs): updated German translation 20081123 12:45:15 mordante: cmake should *not* influence ccache simply due to the way it work 20081123 12:45:17 s 20081123 12:45:28 that is: it just creates the makefiles which use calls to g++ 20081123 12:45:39 and there ccache jumps in since it handles the g++ calls 20081123 12:46:16 well I tried to set the compiler to ccache /usr/bin/g++-4.3 and that didn't work 20081123 12:47:00 uhm, why are you not just having ccache in your PATH? 20081123 12:47:16 that is: the ccache g++ before the "main" g++ in your $PATH 20081123 12:47:56 that is on my system i got this: 20081123 12:47:58 $ which g++ 20081123 12:47:59 /usr/lib/ccache/bin/g++ 20081123 12:48:08 since I might want to change compilers to test some things 20081123 12:48:29 you still can do so 20081123 12:48:48 of course I can fix it but it would be nice if it would work out of the box 20081123 12:48:50 that is: when you want to test different stuff, just insert the complete path to the compiler to use 20081123 12:52:24 mordante: If you don't want to change your workflow, just stay with autotools :) 20081123 12:53:43 loonycyborgw32 I intent to ;) I like that scons seems to make it easier for Windows users and I don't mind to have multiple build systems 20081123 12:54:03 mordante: okay, with ccmake you should easily be able to alter the gcc used: 20081123 12:54:10 just change CMAKE_CXX_COMPILER to the one you want 20081123 12:54:15 fendrin: Please don't hack UtBS in a branch. What's the point? It's just going to cause grief when we have to re-merge. 20081123 12:54:17 Besides, make file.o and make have equivalents with scons too. 20081123 12:54:39 Ivanovic and it than still uses ccache by default? 20081123 12:54:43 esr: the point is to do the "real" work which will alter the campaign, rewrite scenarios and such 20081123 12:54:56 I did some test with some variables and then ccache was no longer used 20081123 12:55:00 mordante: it uses ccache if you point it to the ccache g++ binary 20081123 12:55:23 sure, if you call the "plain" g++, you will not be using ccache 20081123 12:55:35 Why not do that in the main copy? We are, after all, in a development branch. 20081123 12:55:57 esr: when rewriting/adding scenarios strings might change in a really intrusive way 20081123 12:56:49 loonycyborgw32 I assume they do but does vim also know about them, it knows about make 20081123 12:57:28 :set makeprg=scons\ -u or so 20081123 12:57:37 mordante: would be nice to know which vars... 20081123 12:57:52 mordante: that is: cmake has two vars for the compiler, one for gcc, one for g++ 20081123 12:58:05 the CXX vars 20081123 12:58:52 But they're going to have to change anyway when the branch is re-merged! All this does is add another failure poiint to our workflow. 20081123 12:59:26 mordante: uhm, when changing the CXX flags you will have to recompile 20081123 12:59:39 that is: sure ccache will not have the stuff cached anymore when you do so 20081123 12:59:52 esr: yes, but the remerge is planned for *post* 1.6 20081123 12:59:52 I know I can change build systems and fix the scons/cmake issues I have, but like I said; I rather spend my time coding 20081123 13:00:01 And it means my normal wmlscope/wmllint checks won't catch problems introduced in the branch. 20081123 13:00:10 don't fix what ain't broken ;) 20081123 13:00:29 mordante: *and* with the current buildsystem when playing the with cxx flags you have to recompile, too 20081123 13:00:42 but right, as long as autotools work nicely, use it 20081123 13:00:56 esr: correct 20081123 13:01:12 This is a BAD IDEA. 20081123 13:01:53 esr: there are two options: 20081123 13:02:07 1) work on it outside of the main tree so that it is ready post 1.6 and can be taken over 20081123 13:02:14 I know I've to recompile but I have several scripts to do the job 20081123 13:02:18 2) not work on it since changes are too intrusive for 1.6 20081123 13:02:28 wait for 1.6 being released and work on the stuff then 20081123 13:03:09 esr: and yes, in those regards i prefer 1) even though wmllint and wmlscope won't work for some time on it 20081123 13:03:30 though it is possble to just checkout the branches/utbs folder into data/campaigns/utbs 20081123 13:04:00 mordante: Btw latest trunk tries to set data directory to K:/Documents and Settings/Sergey/My Documents/svn-checkouts/wesnoth/K:/Documents and Settings/Sergey/My Documents/svn-checkouts/wesnoth 20081123 13:04:00 How about option 3, not branching the thing until it becomes clear we absolutely need to, or option 4 , not distorting our workflow by making the string freeze more important than actually improving the game? 20081123 13:04:16 With predictable results :) 20081123 13:04:43 * loonycyborgw32 hopes that release revision isn't affected. 20081123 13:04:52 esr: yes, 3 sounds good 20081123 13:04:56 4 is not acceptable 20081123 13:04:57 no I did that change after the release 20081123 13:06:13 esr: especially since eg fendrin_ already said that he will not have *this* much time over the next weeks so it will basically be drop by drop 20081123 13:06:16 I'll have a look at it, but that means wesnoth relative_data_path also doesn't work 20081123 13:06:57 * esr is getting rather angry. 20081123 13:07:32 esr: you should relax a little 20081123 13:08:33 esr: and yes, the "general workflow" we have esteblished since 1.0 is that we got some time for "very intrusive changes" and after a while get closer to a new stable line 20081123 13:08:40 Ivanovic: If that was intended to make me less angry, it didn't work. 20081123 13:08:52 then the main work is basically restricted to bugfixing and polishing 20081123 13:09:07 this is also the time for translators to catch up on the work 20081123 13:09:25 and yes, we are entering the "cleanup and bugfix"-time right now... 20081123 13:10:04 and hey, i remember well that you were asking for a "more reliable" shedule regarding stable releases 20081123 13:10:31 so i think it is a good idea to get a new stable series out close to the start of every new year 20081123 13:10:39 so that the "spring distri updates" can include it 20081123 13:11:15 we could also state that we don't care for stable releases anymore and just release new dev versions whenever we feel like 20081123 13:11:35 though i prefer to offer a stable series once a year where players can "just play" 20081123 13:13:31 esr: have you read the chat log? 20081123 13:14:05 You are dancing around the issue that is making me angry. Once again, I am being told that the process of actually *improving the game* is going to be impeded - in this case by causing tricky merges and making it harder to run my sanity checkers - because you think keeping our translators happy is moreimportant. 20081123 13:15:21 esr: i am not only thinking about making translators happy but also about making non english speaking players happy 20081123 13:15:28 as i already explained several times 20081123 13:15:56 yes, you don't see anyone not speaking english well enough to enjoy the game in it worth to consider in regards to "player base we target" 20081123 13:16:03 i have a different point of view on this 20081123 13:16:17 and i will hold on to this position 20081123 13:16:27 I also prefer stable releases, it also focussed the devs to have a bug fix phase 20081123 13:17:26 also I think that big WML changes should be seen as a feature 20081123 13:18:04 since they often change the balance and I prefer to get some feedback on the balance as well 20081123 13:18:16 yes 20081123 13:18:17 And I don't buy your explanation, for reasons *I* have already explained. We could simply not ship partly translated campaigns. 20081123 13:18:35 Stable releases is a different isue. 20081123 13:20:25 Now I see my attempts to maintain the quality of *the actual game* being increasingly blocked because our translation infrastructure is creaky and our policy decisions related to it actively perverse. 20081123 13:20:40 esr: and if we got string changes up till the last minuite we will have *no* translated campaigns (when leaving out campaigns that are not completely done) 20081123 13:20:52 I think we just see the importance of translations differently and we're not going to convince another 20081123 13:20:57 In my opinion the development releases are mostly for bug fixing. No need for all translations to be perfect. We should aim for perfekt translation only for stable releases. 20081123 13:21:08 fendrin_: jupp 20081123 13:21:35 fendrin_: though eventually translators will have to really work on things which is no fun at all if all the stuff can change all the time 20081123 13:21:35 but we are in front of a stable release right now. 20081123 13:21:37 yes but translators need quite some time to catch up 20081123 13:21:38 If you agree with fendrin, then why in the hell the UtBS branch? 20081123 13:21:56 esr: because of this: [13:21:35] but we are in front of a stable release right now. 20081123 13:23:22 That's perverse. If the translation of development UtBS doesn't have to be perfect, *there's no need to string-freeze it*! 20081123 13:23:46 And thus no need to branch it. 20081123 13:24:21 And thus no need to make my sanity-checking tools ineffective. 20081123 13:26:09 The project's practices don't make any sense even in terms of their own assumptions. That's what's maing me angry enough to consider leaving the project right now. 20081123 13:27:02 esr: they make a lot of sense 20081123 13:27:14 esr: that is when asuming that we are currently preparing for a new stable release 20081123 13:27:34 since up *until* now the state of strings in utbs was not important 20081123 13:27:48 but *now* wo work on getting the new stable series to be stable 20081123 13:27:48 Ivanovic, esr , mordante : let's calm down. No work has been done to this branch right now. So no harm is done. I am going to drink a nice cup of tea and then we discuss the matter in peace. 20081123 13:28:04 fendrin_: i am already drinking a nice cup of tea 20081123 13:28:10 want some earl gray, too? 20081123 13:28:35 I'm having a cup of soup, will make tea afterwards 20081123 13:29:13 I'm getting breakfast. 20081123 13:29:48 * boucman is going to cook his lunch... timezones are fun :) 20081123 13:30:11 :) 20081123 13:31:51 I'm eating breakfast too, which might seem slightly strange since I'm around seven timezones east of esr ;) 20081123 13:33:01 the tea Ivanovic is drinking is his breakfast... 20081123 13:37:19 I'm back. 20081123 13:38:40 * loonycyborgw32 has finished 1.5.6 installer. Multibooting to u/l it to sf. 20081123 13:38:43 -!- loonycyborgw32 [i=kvirc@79.139.247.220] has quit ["KVIrc 3.4.0 Virgo http://www.kvirc.net/"] 20081123 13:40:51 -!- loonybot [n=loonybot@79.139.247.220] has joined #wesnoth-dev 20081123 13:41:09 -!- loonycyborg [n=sergey@79.139.247.220] has joined #wesnoth-dev 20081123 13:43:00 okay, i have sent a mail to the dev-ml regarding the schedule for 1.6 20081123 13:43:05 everyone: please comment on it! 20081123 13:50:59 well, the question to me is why exactly the utbs branch is used. is it because intrusive changes are planned for utbs that will likely break it or are people afraid to do some changes/fixes to it while there is a soft string freeze? 20081123 13:52:01 Soliton: It's the soft string freeze. 20081123 13:52:32 so what kind of changes are you planning that break it? 20081123 13:53:17 Ivanovic I'll comment later, I need to see how fast my wrapping rewrite will be done. I hope to get done today or early next week if not I might not get my todo list done before 14 december :/ 20081123 13:53:42 I think the wrapping isn't too hard, but I had some setbacks before :( 20081123 13:53:52 Some sidemarks to IftU. 20081123 13:54:09 Nothing really important. 20081123 13:54:51 So we thought this can go in a branch until all string freezes are off again 20081123 13:55:00 fendrin_: uhm, shdaow_master already said that *nothing* should be build onto IftU 20081123 13:55:07 The branch is a bad idea for lots of reasons. Two of them are (a) merging sucks. It's a significant failure point, and (b) branches break my wmlscope/wmllint process in a serious way. That hurts the quality of the WML we ship. 20081123 13:55:08 well, then i agree that the branch is silly. 20081123 13:55:10 since he is unlikely to be able to make it work again any time soon 20081123 13:55:43 Ivanovic: doesn't mean you can't reference its story... 20081123 13:56:03 Soliton: oh, you can 20081123 13:56:18 okay, than let's delete this branch again. 20081123 13:56:19 but you should not expect players to have any idea what you are referencing there 20081123 13:56:23 problem solved :-) 20081123 13:56:53 Ivanovic: iftu plays after utbs, you'd rather not know what iftu is referencing now. 20081123 13:56:58 The reason I am seriously pissed off right now is that Ivanovic seems to want not just to put convenience for the translators abovegood new content, but even above the process for making sure our *existing* content doesn't suck. 20081123 13:57:23 esr: Btw have you tried git? 20081123 13:57:47 esr: i think there is a misunderstanding about what the branch is really needed for. 20081123 13:57:48 loonycyborg: Not yet. It's on my to-do list. 20081123 13:57:51 esr: so *where* does our existing mainline content suck? 20081123 13:57:56 loonycyborg: merging in git isn't that painfull? 20081123 13:58:20 can svn merge? 20081123 13:58:36 esr, Ivanovic it seem that you are only disagreeing on what part to freeze and what part not to... 20081123 13:58:45 (no. except very recent versions that need server upgrades afaik.) 20081123 13:59:19 if a few campaigns need to freeze a bit later, that's fine, but we should softfreeze ~60% of game now 20081123 13:59:22 Ivanovic: I keep it from not sucking (at least on one important level) by running wmllint/wmlscope freqently and fixing problems before they snowball. Branches would make that a lot harder to do. 20081123 13:59:50 esr: and we just said that this branch is not needed 20081123 13:59:57 [13:56:18] okay, than let's delete this branch again. 20081123 14:00:09 fendrin_: Yes. Or so I've heard :) 20081123 14:00:20 I missed that wile I was off showeringt, I guess. 20081123 14:00:37 another issue that is important to me and related to the topic is DM 20081123 14:00:47 i am working on DM right now. 20081123 14:00:58 * loonycyborg haven't tried branching and merging with git himself yet. 20081123 14:01:13 fendrin_: this one should IMO really be delayed till post 1.6 20081123 14:01:20 * Dragonking looks at mail regarding wesnoth 1.6 20081123 14:01:32 I do see a good chance to bring DM in mainline before 1.6 20081123 14:01:43 Branch-and-merge is a pain point whatever VCS you're using. The problems are intrinsically hard. 20081123 14:01:55 Will need to find time to look at formula AI and think about what should be improved/fixed sometime soon... 20081123 14:02:02 Half the scenarios are imported from LoW. They are in good state. 20081123 14:02:27 The other scenarios don't seem to be broken either. The wml is looking fine. 20081123 14:02:45 DM strings are stable. I haven't had to do more than typo fixes for weeks. 20081123 14:02:50 well, what's the need to get it into 1.6 though? 20081123 14:02:52 ah cool, a stringfreeze war 20081123 14:03:22 okay, if the DM strings are stable import it to svn head right now. 20081123 14:03:25 i wonder if i should throw in my couple of cents 20081123 14:03:29 I think having mainline reference UMC sounds a bit odd in general 20081123 14:03:40 esr: do you agree with http://www.wesnoth.org/wiki/TextdomainStatus ? (i.e do you consider that status up to date) 20081123 14:03:53 fendrin_: why is it not enough to have it on the addon server? 20081123 14:04:16 mordante: i repeat iftu is after utbs, the references are not what you think they are like. 20081123 14:05:09 Ivanovic: Delfadors Memoirs closes some holes in the history of wesnoth. 20081123 14:05:25 string freezing quite a while before the stable release or branching is fine, it's a reasonable thing to do for translators i guess. what i don't like is freezing when people would still really like to continue making some changes. 20081123 14:05:44 Soliton no I don't know what the references exactly are, but I said it in general 20081123 14:06:18 what subset could we reasonably freeze ? 20081123 14:06:30 Ivanovic: Having it mainline would settle this events in the timeline. 20081123 14:06:53 fendrin_: that's really not a reason to add a campaign short before a stable release... 20081123 14:06:53 bc if we are only disagreeing about utbs, we could mark it for translators to do last and that would be fine 20081123 14:07:06 Ivanovic what's the status of ANL? 20081123 14:07:12 mordante: i have *NO* idea 20081123 14:07:16 there are and will always be holes in the timeline of wesnoth. 20081123 14:07:33 the last stuff i heard was from zookeeper that it is not working though there are no real details about it 20081123 14:08:02 Well, what are valid reasons for including a campaign to mainline? 20081123 14:08:07 Sirp_, did you ever manage to take a proper look into the ANL OOS? 20081123 14:08:14 at this point? none. 20081123 14:08:49 fendrin_: shortly after branching a stable series off: the campaign is good, fun to play and has an active maintainer 20081123 14:09:01 maybe we should have another little ANL test game tonight? 20081123 14:09:05 right now, shortly before the new stable lines: there are no reasons i could think of 20081123 14:09:10 zookeeper: i think so. 20081123 14:09:11 zookeeper: sounds good 20081123 14:09:56 fendrin_: that's a bit frustrating, but we want to avoid a debian like situation, there will be other releases after 1.6 and we have to draw a line somewhere 20081123 14:10:03 all right 20081123 14:10:04 -!- loonybot [n=loonybot@79.139.247.220] has quit [Read error: 104 (Connection reset by peer)] 20081123 14:10:32 Okay, having DM not in mainline for 1.6 isn't a problem for me. 20081123 14:10:42 ok. 20081123 14:11:06 we need more quality add-ons too anyway :p 20081123 14:11:10 jupp 20081123 14:11:44 fendrin_: don't worry, experience shows that good non-official campaigns are the first to get in post release 20081123 14:12:09 and get as much audience, since whenever someone asks for good campaigns, it's always the same that are pointed out 20081123 14:12:19 esr seemed to had hope that this goal could be archieved in time. If it is clear that DM can't reach 1.6 I will continue with tuning LoW and work on DM after 1.6 again. 20081123 14:12:22 * boucman remeber how long it took to have utbs mainlined 20081123 14:12:56 fendrin_: that would be cool 20081123 14:12:58 boucman: Yes, he status page seems up to date, except that UtBS is in somewhat better shape (less likelihood of intrrusive changes) since the page was last updated). 20081123 14:13:15 fendrin_: great, and you can also work on utbs just take a bit care about not doing too intrusive string changes and more polishing. 20081123 14:13:38 esr: feel free to update it whenever you think it should 20081123 14:14:14 -!- Elvish_Pillage2 [n=eli@24-177-39-196.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev 20081123 14:14:34 esr: uhm, "Changes still planned (before 1.6) " for utbs are none 20081123 14:14:47 so as in "the state is stable now (regarding strings)" 20081123 14:15:19 and "since the last edit": This page was last modified 16:21, 22 November 2008. 20081123 14:15:20 ;) 20081123 14:15:47 I think UtBS is sttring-stable, that is unless someone does an actual plot-change on it. OK, since I last saw the page, then. 20081123 14:17:50 My issue about DM is not so much about DM itself. It's that I object - very strongly - to keeping good content out of mainline solely due to translations issues. If we have a reasoned judgment that DM is not good enough, that wouldn't bother me - but I don't think translation burden is a good enough reason. 20081123 14:18:47 well, translation burden is not all the reason IMO. it's mostly that we approach a stable release and there is no need to risk a lot of trouble by introducing a new campaign now. 20081123 14:19:17 Soliton: What other sort of trouble do we risk, in your view? 20081123 14:19:54 I am used to having many open source x.0 release only in some or even only in one language. The translations are going to flow in with x.1 and x.2 . This isn't the wesnoth policy? 20081123 14:19:55 it taking a lot of effort to bring up to mainline quality taking developers time off of other issues. 20081123 14:20:38 fendrin_: well, for wesnoth the thing is we had a very high number of translation at version 1.0 and never again so far. 20081123 14:20:44 Am moving my laptop to continue this over breakfast. Connection might drop, have patience... 20081123 14:20:51 complete translations that is. 20081123 14:21:33 Soliton: not so much effort any more, I automated away a lot of it. 20081123 14:22:46 right, every wmllint clean add-on is basically great already. 20081123 14:23:06 fendrin_: most OSS projects have very few strings compared to wesnoth and most of them don't change from one release to the next 20081123 14:23:27 in our case we can't have translations not be in a release because they would never really get in 20081123 14:23:46 basically we are probably the biggest gettextuser out there 20081123 14:23:52 Perhaps there is a technicalo solution. 20081123 14:23:54 (bigger than eg kde) 20081123 14:24:28 Ivanovic: that big ? I never thought so... 20081123 14:24:55 even big programs probably don't have that many strings in their interface. 20081123 14:25:11 and some of our strings are HUGE I know :P 20081123 14:25:13 boucman: we got *many* strings due to the campaigns and such 20081123 14:25:19 * boucman gave a hand to the french translation at some point 20081123 14:25:39 okay, in 1.0 i would have said that kde is bigger 20081123 14:25:52 by now i guess that we are slightly bigger in regards to "amount of work to translate it" 20081123 14:26:20 -!- loonybot [n=loonybot@79.139.247.220] has joined #wesnoth-dev 20081123 14:26:52 esr: the other thing that i disagree with is that it's a big deal to have one less nice campaign shipped with mainline instead of available on the addon server. especially since it's not like we ill never have it in mainline just a little later. 20081123 14:27:16 I agree with all that was said. What I don't understand is: Why can't the translation take place after a stable release. The strings are very frozen at that time. And the translators do have all the time of the world to translate it. 20081123 14:27:42 since that the first release ships with a lot of incomplete strings 20081123 14:27:49 fendrin_: that's a lesson we learnt the hard way in 1.2 vs 1.4 20081123 14:27:54 fendrin_: that's what happens but it'd be nice to have done much of the work already once the stable release gets out. 20081123 14:28:08 * esr thinks fendrin's suggestion is simple and sensible. 20081123 14:28:15 and most translation teams are still catching up with the 1.4 changes 20081123 14:28:24 being ready for stable is a high motivation factor for translators... if it's not translated in 1.x.0, it will most likely never be translated 20081123 14:28:52 and being the first translation to reach 100% and have the original release announcement in your language is anothe highly motivational factor 20081123 14:29:06 so we have to make sure translators have a chance to reach 100% 20081123 14:29:36 maybe we should thing of a translator motivation system that works quite differntly. 20081123 14:29:48 cookies? 20081123 14:29:49 s/thing of/think about 20081123 14:30:01 I'm thinking out load, now. A lot of the conflict here is because translation has so much process friction. Suppose we could reduce the process friction, the way I reduced the process friction on camoaign maintainace with wmllint/wmlscope? 20081123 14:30:13 -!- Elvish_Pillager [n=eli@24-177-39-196.dhcp.oxfr.ma.charter.com] has quit [Read error: 110 (Connection timed out)] 20081123 14:30:14 fendrin_: maybe, but that's the best we have so far... 20081123 14:30:20 IMO opinion shipping first and then let them catch up, makes them feel like second class 20081123 14:31:01 mordante: Not if we present it as a way to save them grief from unstable strings ;-) 20081123 14:31:02 esr I think the easiest fix is to allow new content only early in the dev cycle 20081123 14:31:18 esr: that could help... I think the most important thing would be a way to mark a string change as "significant" or "non-significant" so translators don't have to proofread whenever the original string has a typo fix 20081123 14:31:20 esr: What if we'll automate translation itself with a translation program? :P 20081123 14:31:25 mordante: And that is *exactly* the assumption I object to. 20081123 14:31:46 esr: well, what do you propose? 20081123 14:31:51 loonycyborg: No, but I'm trying to imagine tools to lower the friction. 20081123 14:32:11 OK, I'm imagining now. 20081123 14:32:52 Add babelfish support to wmllint :) 20081123 14:33:00 feel free to think out loud, we are quite few here with experience with translations that might help 20081123 14:33:19 frankly i don't see any problem with not mainlining DM until after 1.6. sure, 1.6 will have one less campaign, but it already has 14...any new players won't be bothered, and old players will know to get it from the add-on server. 20081123 14:33:27 loonycyborg: actually, using it as a spellchecker/vocabulary consistancy checker would be cool 20081123 14:33:43 pinpointing if a unit name is translated differently in different places for instance 20081123 14:34:03 okay, created a page in the wiki: http://www.wesnoth.org/wiki/Wesnoth1.6ReleasePlan 20081123 14:34:11 well I think it would at least help if we state the intention to mainline it and have it on wescamp so people can start on it 20081123 14:34:12 Each campaign has a webpage that's a CGI script. To update atranslation, you go to a page, log in, and look at it. Next to each string is any existing translation, an autogenerated Google Translations rendering if there is no existing translation, and a text form box where you can enter a new one. 20081123 14:34:33 in fact we already said it should be on wescamp before considering it to be mainlined 20081123 14:34:45 boucman: You are behind, wmllint does spellchecking now :-) 20081123 14:34:46 boucman: wmllint already has support for spellchecking. 20081123 14:35:05 esr: err, so you want to improve the tools translators use? 20081123 14:35:18 Adding support for machine translation would be logical next step :P 20081123 14:35:29 esr: multilingual ??? that's great 20081123 14:35:41 well then machines first need to learn to translate properly :P 20081123 14:35:42 Soliton: If that enables us not to have a content policy that I think is *perverse*, yes. 20081123 14:35:53 -!- VladimirSlavik [n=chatzill@136.219.broadband6.iol.cz] has joined #wesnoth-dev 20081123 14:35:54 esr: i don't so how. 20081123 14:36:25 Soliton: Could not understand your last. 20081123 14:36:46 esr: i don't so how that helps our content policy. 20081123 14:37:04 Soliton: I'll explain... 20081123 14:37:07 there is no problem of translations getting done afaict. 20081123 14:37:16 esr: uhm, the stuff with "webpage for it" is not too well suited for a game like wesnoth 20081123 14:37:30 that is it kills consistency in the strings and tends to lower quality 20081123 14:37:48 and yes, that approach is not useful anyhow. 20081123 14:37:51 because then translators might jump in for two or three strings which will not be consistent in style for the campaign 20081123 14:37:54 Soliton: If there's no problem with it, then what *is* the problrem. 20081123 14:37:57 ? 20081123 14:38:02 esr: you tell me. 20081123 14:38:05 sure, for "interface only" programs this works really nicely 20081123 14:38:22 for a game such a "webinterface, enter your stuff, be done" does not work well 20081123 14:38:26 i don't have a problem with our content policy. 20081123 14:38:56 in general for translators the po files we offer for download from out "webinterface" (gettext.wesnoth.org!) if just fine 20081123 14:39:26 after working on the file, it is sent to the maintainer of the lang and they take care of the "quality control" however they think it be reasonable 20081123 14:39:36 bbl bye 20081123 14:39:46 esr: do you want to speed up translations in general? is that the friction you were talking about? 20081123 14:39:52 Ivanovic: I hear your objection, but I think it might be met with some good interface engineering. For example, we could take care to present the strings in the order they occur in the text. 20081123 14:40:00 Soliton: Yes, exactly. 20081123 14:40:13 esr: uhm, they *are* in order in our po files 20081123 14:40:22 esr: that is: we can the files by filename 20081123 14:40:26 esr: magic? 20081123 14:40:30 and the scenarios got those lovely numbers 20081123 14:40:47 esr: a webinterface to the po files is not real help 20081123 14:41:01 what does help is some script to do "real magic" on typo and grammar fixes 20081123 14:41:24 so that "this is a typo/grammar fix only" changes in the english version are automatically propagated to the po filse 20081123 14:41:40 without putting the burdon of "really check the string" on translators 20081123 14:41:44 OK, if you want consistency of style, I can thinki of ways to play with the interface that would channel translators in that direction. 20081123 14:41:44 yes, that is something i would have complained about if i'd be a translator. 20081123 14:41:53 but it seems they don't really care. 20081123 14:42:22 esr: do not reinvent the wheel 20081123 14:42:25 esr: you should look at the existing tools translators use first before imagining what might help them. 20081123 14:42:29 there are good tools for working on the po files 20081123 14:42:41 they work great, even got spellchecking 20081123 14:43:51 For example: We could maintain not one database of translations per file, but one per file/translator combination. You log in, you see and update your translations, and the language maintainer chooses which translator's version to actually ship. 20081123 14:43:54 either way i don't understand what speeding up translations has to do with process friction. 20081123 14:44:03 don't translators currently need to mail their files around to the translation maintainer, who mails them to ivanovic who commits them? if anything, eliminating the need for mailing files around constantly might be nice.. 20081123 14:44:41 Soliton: Because if transltors can update faster and with less effort, the pressure on us to stringfreeze would decrease. 20081123 14:45:03 zookeeper: what I receive as the middle man sometimes makes me want to cry, so eliminating me would definitely lead to lower quality 20081123 14:45:10 esr: nevertheless that has nothing to do with process friction... 20081123 14:45:21 zookeeper points out annother place where we could gain a lot of effficiency. 20081123 14:45:39 VladimirSlavik, right ;) 20081123 14:45:45 well, i would suggest to listen to actual translators.... 20081123 14:46:06 hehe, just wait till i catch up with the whole discussion... 20081123 14:46:25 Soliton: The lag in translations produce process friction for the *entire project*? What do you suppose I have been complaaining about? 20081123 14:46:30 VladimirSlavik, well, my point was that you'd still get to review the changes before forwarding or committing them, just that the interface for doing all that would be more convenient. if it's inconvenient atm, that is. 20081123 14:46:41 First ? in above should be ! 20081123 14:47:48 esr: i simply object to term used since it is misleading, nevermind. 20081123 14:48:35 esr but what already might help is to get the addon on wescamp when it's polished and say we serious consider to mainline it 20081123 14:48:57 that way if translators want to work on UMC they know what has the biggest odds of getting mainline 20081123 14:49:52 mordante, fendrin: Agreed. fendrin, it is (effectively) your campaign; would you get DM up on Wescamp,please? 20081123 14:50:23 Can't the game itself be a translation tool? When the translation option is enabled every message dialog would get a button which opens a text input field for the translation to the language the player plays the game with. This string is send to a maintainer of that langueage who can choose the best one. 20081123 14:50:58 and I think it would be good idea to also list this in the translators forum 20081123 14:51:11 fendrin_: that would lead to horrible inconsistency 20081123 14:51:18 fendrin_: Cool! Now *that* is thinking outside the box. 20081123 14:51:51 getting random people involved in translations is probably not helping quality. 20081123 14:52:01 to be honest for initial translation 1 translator gets much better results as a group 20081123 14:52:05 if we wanted quantity we'd use an automatic translation tool. 20081123 14:52:50 making players more aware that they could simply help in translations might be nice though. 20081123 14:52:53 Ah bvut random people with a filter -- the responsible language maintainer -- that could work. 20081123 14:52:58 mainly to get the story and style consistent 20081123 14:53:06 Probably if really many people would be involved then there would be enough quality strings. 20081123 14:53:21 But translation maintainer would be swamped. 20081123 14:53:26 exactly. 20081123 14:53:55 one could filter out duplicate translations.. maybe. 20081123 14:54:12 but i doubt it'd be fun for the maintainer and he'd rather translate it himself. 20081123 14:54:38 uhm, once again and for last time from me: quality is not per-line 20081123 14:54:39 check 5-10 submitted strings and determine which one is best and even then maybe make some manual tweak to it, or just translate it yourself? 20081123 14:55:17 and yes, i'd seriously fear for quality if, say, the dialogue in one campaign was translated by 5 different people instead of 1. 20081123 14:55:27 IIRC Debian already uses such a webbased system... and I think Rhonda mentioned the quality isn't that great 20081123 14:55:29 since they're not all going to write in the same style and so on 20081123 14:56:44 of course they can currently arrange the translation like that too, but encouraging it doesn't sound like a grand idea 20081123 14:59:37 -!- crimson_penguin [n=ben@64.201.60.217] has joined #wesnoth-dev 20081123 14:59:38 I play the campaigns in english. I'm not that good with english but it's enough for getting the story. But there are people that can't speak english. For them having a bad translation that is good enough to get the point is more important than getting one that has the best quality and is very consistent in a far distant future. 20081123 15:00:21 i would think that depends highly on the preference of said player. 20081123 15:00:22 I disagree 20081123 15:00:30 that's highly speculative. 20081123 15:00:45 if I read badly translated stuff I think the game sucks 20081123 15:00:58 VladimirSlavik: At what granularity do you think 'quality' become meaningful? Continuous stretches of dialog? Scenarios? Entre campaigns? 20081123 15:01:10 if you don't care about the story and just want to know what's going on you're right.. 20081123 15:01:16 in that case we can also lower the bar for art so we have portraits for all units 20081123 15:01:21 If I can't play a game because it has chinese signs it sucks for me as well :-) 20081123 15:02:13 and I'm quite sure most/all native english speakers also hate it when the english sucks 20081123 15:02:41 so if a campaign (where you play a story) has broken text that is maybe still hard to understand for you that will make you like that game? 20081123 15:03:03 mordante: A lot of English speakers are, alas, rather blind to style. 20081123 15:03:20 not only style but grammar as well 20081123 15:03:23 whil writing or reading? 20081123 15:03:27 whileÜ 20081123 15:03:30 hrm. 20081123 15:03:34 esr: that's hard to define, you have more than one "dimension" across which you can achieve it 20081123 15:03:47 nice examples of errors Soliton :P 20081123 15:03:52 esr: generally I prefer whole campaigns 20081123 15:03:55 :-P 20081123 15:04:05 VladimirSlavik: I'm interested, and listening, and it's relevant to the larger questions. 20081123 15:04:48 maybe I should write down what I think and post it somewhere rather that chop off small pieces here? 20081123 15:04:53 http://www.wesnoth.org/forum/viewtopic.php?f=4&t=23043 <-- what is the screen size of eee pc? 20081123 15:05:05 VladimirSlavik: Post on the ML. 20081123 15:05:12 which one? 20081123 15:05:41 Dev list - this relates to a policy issue for the whole projecct. 20081123 15:05:44 iirc 800x480, need to use the --small-gui start parameter to be able to use that 20081123 15:05:45 ok 20081123 15:05:56 loonycyborg: my eee pc has a 10" display with 1024x600 20081123 15:06:42 -!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["Lost terminal"] 20081123 15:08:08 mordante: It's --smallgui :) 20081123 15:08:41 hmm some people hate hypens ;) 20081123 15:12:31 -!- ilor_ [n=a@unaffiliated/ilor] has joined #wesnoth-dev 20081123 15:13:46 -!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev 20081123 15:16:37 -!- ilor [n=a@unaffiliated/ilor] has quit [Read error: 60 (Operation timed out)] 20081123 15:20:04 -!- BenUrban [n=benurban@unaffiliated/benurban] has quit [Read error: 104 (Connection reset by peer)] 20081123 15:20:21 -!- BenUrban_ [n=benurban@unaffiliated/benurban] has joined #wesnoth-dev 20081123 15:29:20 -!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["Lost terminal"] 20081123 15:36:39 esr: in general i think the basic process on the translators side we got (leave me some mins to completely describe it!) is rather good 20081123 15:36:41 that is: 20081123 15:37:05 1) we got a translation maintainer feeling responsible for the translation as a whole and its quality 20081123 15:37:26 -!- esr [n=chatzill@146.145.247.101] has quit [Read error: 145 (Connection timed out)] 20081123 15:37:27 2) we got normal translators who do translate the stuff and send it to their maintainer 20081123 15:37:37 (i hope he reads the logs) 20081123 15:37:58 this maintainer will then make sure that the quality is okay for him to have it included in the game 20081123 15:38:17 at the beginning this is probably done by reading over the changes and correcting them 20081123 15:38:36 then he will forward this to torangan or me for inclusion in the game 20081123 15:39:20 after a while the maintainer gets to know the people who work with him on the stuff as in "hmm, the work from XYZ can directly be forwarded, the one by ABC has to be corrected to be of worth" 20081123 15:39:26 this way the maintainer acts 20081123 15:39:55 this way he is able to make sure that the quality of "his" translation is what he thinks is good enough 20081123 15:40:05 basically this works out rather well 20081123 15:40:18 what can be improved on the other hand is the *unneeded* maintaince overhead 20081123 15:40:39 which is created by changes to strings in the english version that do not alter anything important (like fixing a typo) 20081123 15:41:08 currently this string is marked as "has to be fixed" and the translators got no way to see if it is a big change or just a "yes, this is fine anyway..." 20081123 15:41:19 afk for ~30mins, lunch... 20081123 15:54:19 -!- crimson_penguin [n=ben@64.201.60.217] has quit ["Leaving"] 20081123 15:56:03 Ivanovic: Windows installer for 1.5.6 is uploaded. 20081123 16:00:58 -!- esr [n=chatzill@146.145.247.101] has joined #wesnoth-dev 20081123 16:04:22 re 20081123 16:04:26 loonycyborg: *great* 20081123 16:04:37 esr: i left a little in the logs for you regarding the normal workflow 20081123 16:04:41 with translations 20081123 16:05:17 that is basically: 20081123 16:05:39 the biggest burden that there is currently are those tiny fixes we got in mainline 20081123 16:06:11 fixes that only fix typos and grammar mistakes without changing the meaning or anything of importance (newlines are important as well as the change of a . to a :) 20081123 16:06:30 what gettext is basically doing when running an update on the files is: 20081123 16:07:03 it takes the catalog (eg wesnoth.pot) and looks for all strings listed in this file in the "to be merged" translation file 20081123 16:07:24 if the english string is exactly the same way in there, it is just taken over (see it like an md5sum on the string) 20081123 16:07:46 if the *exact* string is not found, it searches for the "nearest neighbour" 20081123 16:07:57 so if it finds a version that is identical beside one char, this one is taken 20081123 16:08:25 since it was not an exact match, this string is marked as 'fuzzy' which basically means "hey, translators, i guessed this string and it might fit, you got to check it" 20081123 16:08:42 and if you now consider that we got a huge campaign like utbs 20081123 16:09:00 we fix a huge load of typos and grammar effecting about 100 strings (just example numbers) 20081123 16:09:07 those 100 changes mean 100 fuzzy strings 20081123 16:09:35 but we also got 5 *real* changes where the meaning somehow is switched (north becomes south, introduction of place names, ...) 20081123 16:09:48 so at the end we got 105 fuzzy strings 20081123 16:10:08 where the translators now got to sort out which ones are fine and are already okay and which ones are not and have to be corrected 20081123 16:10:42 ingame only strings are shown which are a) translated and b) not marked fuzzy 20081123 16:11:15 if we now had a good way to remove the burden of the 100 "no work to be done" fuzzy strings, this would be a real improvement in the workflow 20081123 16:11:25 the problem is that this is a non trival problem 20081123 16:13:07 on the other hand a webinterface just to the po files will most likely not offer anything special to translators which actually helps improving the translation 20081123 16:13:28 since the tools to work on po files are really good by now and even got good spellcheckin included 20081123 16:13:57 VladimirSlavik: have i missed anything in this description of workflow and what would help translators most? 20081123 16:14:31 Ivanovic: I have one idea for that. In case of a small typo fix, would it be possible to have some sort of script go through the po files and update the source string to the changed version? 20081123 16:14:50 ilor_: in theory this would be the possible solution 20081123 16:14:53 Sounds like one good step would be to remove all punctuation before doing fuzziness checks. 20081123 16:15:03 ilor_: just some problems with this: 20081123 16:15:14 1) strings can be in a files twice and almost identical 20081123 16:15:20 esr: sometimes the punctuation can be important. I'd be careful with that 20081123 16:15:25 so with the replacement you might replace something you don't want to replace 20081123 16:15:25 punctuation changes might need a new translation as well 20081123 16:15:48 2) we need a *good* list of those strings which is 100% correct for the string 20081123 16:16:00 esr: uhm, punctionation changes are a nice problem 20081123 16:16:08 Ivanovic: the script should be smart enough to only replace the one surce-string it actually is 20081123 16:16:16 that is: what to do about this string: the string "team" is ingame 20081123 16:16:29 the string "team:" is added and used at a completely different place 20081123 16:16:49 of course gettext sees that team: is reasonably close to team 20081123 16:16:55 so it choses it 20081123 16:17:18 still from an interface point of view the : there might make all of a difference 20081123 16:17:38 ilor_: we would need a skript where the *complete* string to be replaces is added to 20081123 16:18:00 ilor_: and the script would have to transform the po file to "one line for the english string" before doing the matching 20081123 16:18:28 currently strings in po files can easily go over several lines since the progs tend to introduce a newline every 80chars 20081123 16:18:39 this all seems doable 20081123 16:18:44 yes, it is 20081123 16:18:56 the main problem is just maintaining this "database of changes" 20081123 16:19:10 I don't think there should be any database 20081123 16:19:22 ilor_: even if it is just a plain list it would be needed 20081123 16:19:37 and no, we can not run a pot-update for every single fixed typo 20081123 16:19:40 the script should be capable of taking a svn diff where a typo is fixed in a source string 20081123 16:19:59 (which would be needed after changing the "source string" in each po file) 20081123 16:20:25 I don't exactly know how pot-update works 20081123 16:20:27 ilor_: the strings representation in po files and in wml files is a little different 20081123 16:20:36 a pot update is basically this: 20081123 16:20:57 * with some magic a list of current strings ingame is collected and written to textdomain.pot 20081123 16:21:14 * all po files in this textdomain are merged against this new textdomain.pot file 20081123 16:22:06 -!- AI0867 [n=ai@van-gessel.demon.nl] has joined #wesnoth-dev 20081123 16:22:38 the "magic" part being that a list of files found with the command in FINDCFG is passed over to wmlxgettext which extracts the strings from the config file and gettext tools are searching the cpp files which are listed in the respective POTFILES.in for strings, at the end the resulting two lists are just put together as one list 20081123 16:24:18 Ivanovic: what would happen if the msgid in a .po file would be changed before a pot-update? 20081123 16:24:32 in general it would be possible to add a call to a script that transforms the existing files using the list of "english strings that have not changed" before running the 2nd step of the potupdate 20081123 16:24:52 that is: then the matching part in the msgmerge would see "hey, the string is there, i can take it" 20081123 16:25:06 it is just that we do need a convenient way to do the matching 20081123 16:25:42 in a scenario like this: there's a "good mrning" string, translated to "guten morgen" in de.po 20081123 16:25:50 which will probably require to transform the po files in a way that the english messages (called 'msgid') are in a and single line per string 20081123 16:25:59 the typo gets fixed, so the string changes to "good morning" 20081123 16:26:15 the de.po file gets updated by our magic yet-to-be-written script 20081123 16:26:34 and the msgid there changes to "good morning" with the translation unchanged 20081123 16:26:42 afterwards in the pot-update this would not be visible that something changed 20081123 16:26:54 so no fuzzy string 20081123 16:26:57 yes, this would be possible 20081123 16:27:10 and in a perfect world it would work out *perfectly* 20081123 16:27:21 now comes on of the other problems: 20081123 16:27:37 translators work on files for some time before sending them in to have them updated against the latest catalog 20081123 16:28:07 so if we do fix the strings silently and then the translator send his (now outdated) file in, those strings will get marked fuzzy 20081123 16:28:07 *That* is the issue a we b interface would address directly. 20081123 16:28:50 esr: i have two issues with the webinterface idea (like i already said above, too): 20081123 16:28:58 1) reduced quality 20081123 16:29:11 Ivanovic: ah, that's where the database of fixes idea came from 20081123 16:29:16 2) there are already great tools for po files, why not use them 20081123 16:29:19 ilor_: exactly 20081123 16:29:33 ilor_: no need for a "real" database kind thing, just some pseudo system 20081123 16:29:49 Ivanovic: it's a database even if it's a flat file ;) 20081123 16:30:05 I wonder how well svn merge would work 20081123 16:30:22 ilor_: yeah, just wanted to say with this that no "full fledge sql or whatever database" is needed 20081123 16:30:28 ilor_: not good at all 20081123 16:30:43 Ivanovic: yeah probably 20081123 16:30:43 since the strings are updated against the .pot file before uploading 20081123 16:30:57 Ivanovic: If the result of the "great tools" is a workflow that is laggy, friction-filled, and has effects on the rest of the project that piss off someone as senior as me, then the tools ain't so great. 20081123 16:31:06 we (as internationalization managers) do make sure that the files are up to date 20081123 16:31:22 esr: and for a webinterface you got to be online all the time 20081123 16:31:34 sure, here in the western world this is no problem at all 20081123 16:31:38 I think we could use svn anyway to track these typo fixes 20081123 16:31:51 but there are places in the world where there is not this much internet 20081123 16:32:09 ilor_: difficult, like i said 20081123 16:32:27 or you would have to eg seperate your commits into parts that are "typo fix only" and other that change meaning 20081123 16:32:30 Ivanovic: I don't think it's that difficult 20081123 16:32:38 and eventually someone will make a mistake 20081123 16:32:48 Ivanovic: yes that's one thing that I think would have to be enforced 20081123 16:32:58 and commit a "real" change, too 20081123 16:33:27 that's always a risk 20081123 16:33:41 esr: so what about people with not this much internet access who eg download the po file on an usbstick, work on it at home and send it in a week later when they got internet again? 20081123 16:33:54 those can not work on it anymore when we got such a webinterface 20081123 16:34:17 and yes, not everyone on this world has a flatrate so those webtools can be used 20081123 16:34:56 ilor_: that is why we would probably need a system where manually is entered "this string change is just a typo fix" 20081123 16:35:09 ilor_: best would probably be if this system was used directly to fix the original string 20081123 16:35:39 Ivanovic: I'm thinking of how this could work 20081123 16:36:12 then this would be added to the "database" as fix to automatically correct and be done so when running the update commands 20081123 16:36:17 esr: how about translations take time? i still don't see what you mean with "friction-filled"... 20081123 16:36:23 Ivanovic: There are always tradeoffs. We'd lose some of those people, yes, but we'd also get rid of a lot of duplication and lag. Unless you think we have huge player consituencies in countries with very poor internet, the tradeoff is in favor of a web interface -- and bound to tilt more that way as the net infrastructure builds out further. 20081123 16:36:38 yes, probably some typo fixes would be done in a different way and thus be marked as fuzzy though not requiring work 20081123 16:36:48 but i prefer false positives to false negatives 20081123 16:37:01 esr: we would not get rid of much of the lag 20081123 16:37:07 esr: translations need time 20081123 16:37:16 as well as writing good prose needs time 20081123 16:37:23 and translations basically are the same 20081123 16:37:37 Soliton: I mean this process where translations go out in whole-file form only to return weeks later (if at all) suring which time the underlying string base might have changed. 20081123 16:37:54 esr: okay, what if i wanted to work on things, download the po file on my laptop to work on it in the train 20081123 16:38:02 this is not possible anymore with a webinterface 20081123 16:38:15 sorry, webinterfaces are *NOT* the one tool fits all solution 20081123 16:38:25 * esr has web on his cellphone. 20081123 16:38:38 esr: in germany internet "mobile" is damn expensive 20081123 16:39:05 and i know not many people who do use it (that is i know exactly one) even though i study computer science with lots of geeks around 20081123 16:39:14 * esr has an Android phone from former Deaursche Telekom :-) 20081123 16:39:20 yes, with the cellphones you got the *option* to use it 20081123 16:39:24 * AI0867 is planning to buy one of those 20081123 16:39:29 but they charge a real fortune for doing it 20081123 16:39:36 s/Deaursche/Deutsche/ 20081123 16:39:53 but yes, even in the most connected areas of the world, having internet *everywhere* isn't common 20081123 16:40:08 The only problem I see with translation is the time they take place in. We do a string freeze. The translaters do their work. After that a release is done. At least strings can be manipulatet again. I would try to motivate the translaters to do their magic after the release. Just call the x.0 a translaters release special for them. They feel honoured and are motivated as hell. x.1 is then called the translated final. 20081123 16:40:09 (where i know many people who got mobile phones that are able to eg use 3G, it is just not affordable to really do use it) 20081123 16:40:42 if i read the backlog do i find anywhere a description of said web interface? 20081123 16:40:49 Ivanovic: But this is a red herring. The right thing would be to have both interfaces and see which one yields well. 20081123 16:40:55 fendrin_: and many distributions (at least the debian based ones) only include the version that is available when they release 20081123 16:41:02 those never get the translation updates in 20081123 16:41:04 Soliton: Only a speculation of mine. 20081123 16:41:27 so translators should at least have a chance on having a lot ready with the initial release 20081123 16:41:49 yes, this takes a while and we have to restrain from edits for this time 20081123 16:42:54 esr: like i said: i would prefer to have a nice way to update the po files when they are sent in, too 20081123 16:42:59 also, when is the translated final release? 20081123 16:43:11 (that's right, never.) 20081123 16:43:44 you have to have an arbitrary cut-off point just as we have now. 20081123 16:43:45 Ivanovic: You don't have a po merging script? 20081123 16:43:48 a way which does only not mark strings as fuzzy that really are not changed in a meaningfull way 20081123 16:44:13 esr: just assume that computers are damn stupid and know nothing for real about english language 20081123 16:44:35 so i can't tell any "plain script" something like "please search for spelling and grammar fixes only and do not make those fuzzy" 20081123 16:44:42 such a general command is so far not possible 20081123 16:44:51 we at least have to gather this list ourselfes 20081123 16:44:55 and this in a somehow nice way 20081123 16:45:00 I understand that. 20081123 16:45:11 yes, merging of files on a strict "the id is identical" is really no way and dead easy 20081123 16:45:26 s/no way/in no way a problem 20081123 16:47:40 and the normal merging of files is done by gettext via the tool "msgmerge" 20081123 16:48:10 i have described how it works one hour ago 20081123 16:48:24 though you left right when i started explaining it, so please read it up in the logs 20081123 16:48:31 loonycyborg * r31008 /trunk/src/game.cpp: Fixed incorrect detection of absolute path on windows. 20081123 16:48:32 http://wesnoth.org/irclogs 20081123 16:48:47 afaiui what would be good to have for managing po files is: a) a script to intelligently unfuzzy strings in all translations for typo fixes. b) fix the current tools to not always rewrite the whole po file when translating. c) allow/fix that --previous parameter or whatever to let translators see what exactly changed. 20081123 16:49:04 I still think committing a commit with *only* spell fixes with a special marker would be the best way and then write some magic stript that does the pot-update and po merges 20081123 16:49:26 b) is so you can use patches and not have to overwrite complete files and wipe out any intermediate changes. 20081123 16:49:35 mordante: Be sure to review r31008 20081123 16:49:45 If you guys can agree on a specification, I can write the tool. 20081123 16:49:46 already looking at it ;) 20081123 16:50:10 esr: the specification is the main problem 20081123 16:50:22 it is difficult to exactly state the stuff that has to happen 20081123 16:50:30 uh? 20081123 16:50:44 mordante: this is problematic and might lead to false positives 20081123 16:51:04 Soliton: the exact steps to make sure that nothing wrong will be changed 20081123 16:51:15 i don't see the problem. 20081123 16:51:18 Soliton: since we need to know a) the textdomain to work on 20081123 16:51:23 loonycyborg it's tricky but should work 20081123 16:51:29 bbl... 20081123 16:51:34 (the "old" text could be in several though only in one it has to be changed) 20081123 16:51:55 b) extract the *exact* string how it will appear in the po file 20081123 16:52:18 at least I need to check but I think it's save 20081123 16:52:19 (including rather simple stuff like transforming newlines into explicit \n) 20081123 16:53:07 the problem basically with this specification is that i have no idea about all the corner cases and which ones there might be 20081123 16:53:15 there are probably several that you don't easily think about 20081123 16:54:13 "so let's not even try" 20081123 16:54:40 :) 20081123 16:55:09 Rhonda: in which gettext version was the --previous command introduced? 20081123 16:55:28 loonycyborg: which version of gettext are you using to create the translation messages for windows? 20081123 16:57:41 loonycyborg yes it should work, I'll add some comment and also fix the case if the user gives a parameter 20081123 16:59:17 mordante * r31009 /trunk/src/display.cpp: Remove some headers already included in display.hpp. 20081123 17:02:27 Espreon: Eloh Cultisys <-- this team name is used in a lot of places, it looks like a typo 20081123 17:03:20 Aaargh! I don't spellcheck team names, or I'd have caught that. 20081123 17:04:00 esr: the player teams in LoW are called kalenz 20081123 17:04:10 difficult to spellcheck 20081123 17:05:02 No it isn't -- I already have him declared a local spelling exception. 20081123 17:09:29 esr: do you spellckeck the names of variables? 20081123 17:10:40 ai0867 * r31010 /trunk/data/campaigns/Legend_of_Wesmere/_main.cfg: Undef a macro. 20081123 17:10:53 fendrin_: about the LoW about.cfg: the long line about the community isn't linewrapped, so it just falls off the side 20081123 17:11:59 AI0867: wasn't about.cfg merged to main.cfg because a automatic tool couldn't handle external abouts? 20081123 17:12:06 mordante * r31011 /trunk/src/game.cpp: Also port r31008 to the startparameters and add some comment. 20081123 17:12:07 loonycyborg small addition^ 20081123 17:12:43 fendrin_: er, yeah 20081123 17:12:52 but the issue is still there 20081123 17:16:37 AI0867: just adding a newline doesn't solve the problem? 20081123 17:24:41 mordante * r31012 /trunk/src/gui/widgets/debug.cpp: Truncate labels langer as 50 chars. 20081123 17:24:46 mordante * r31013 /trunk/src/gui/widgets/debug.cpp: Fix a bug where the same string was alsways returned. 20081123 17:26:23 Ivanovic: msgfmt.exe (GNU gettext-tools) 0.16.1 20081123 17:26:27 Going offline. Travel imminent. 20081123 17:26:31 -!- esr [n=chatzill@146.145.247.101] has quit ["ChatZilla 0.9.84 [Firefox 3.0.4/2008111318]"] 20081123 17:26:53 loonycyborg: okay, then for windows it would be no problem if i switched gettext min version required to 0.16.1 20081123 17:27:03 would like to know about OSX, too 20081123 17:27:33 and then maybe make the min version of gettext for the game 0.16.1 (or whenever --previous was introduced to msgmerge) 20081123 17:28:04 since by now i hope that all the tools support it nicely 20081123 17:28:30 (though debian stable still has not updated the version of poedit that is shipped to one that works with it, but who cares about debian anyway? 20081123 17:28:32 ) 20081123 17:30:25 Ivanovic: iirc the problem was last time that gettext for mac was outdated 20081123 17:30:37 VladimirSlavik: yes, mac was a problem 20081123 17:30:39 and poedit, too 20081123 17:31:04 that is why i wanted to know from crimson_penguin which version of the tools he used 20081123 17:31:09 I am rather disappointed by poedit 20081123 17:31:22 VladimirSlavik: it works nicely for me 20081123 17:31:23 ;) 20081123 17:31:43 the author is lacking any incentive to improve it seriously 20081123 17:32:39 fendrin * r31014 /trunk/data/core/about.cfg: added santi and his sons to the about file to give them credit as campaign designers of LoW 20081123 17:32:41 nothing at all happens with the code except smal bugfixes 20081123 17:32:56 VladimirSlavik: right 20081123 17:33:32 and on its mailing list, author asks if people would be willing to pay for commercial tm support 20081123 17:33:39 :-/ 20081123 17:34:45 i don't follow the lists 20081123 17:35:02 that is: i just use poedit in the latest version that is available in gentoo and it works for me the way i need it to 20081123 17:35:13 yes, it is a great tool 20081123 17:35:23 but its future looks rather bleak 20081123 17:35:35 this might be 20081123 17:36:40 fendrin * r31015 /trunk/data/campaigns/Legend_of_Wesmere/_main.cfg: added newlines to too long entries in the abouts of LoW 20081123 17:37:13 automatic newlines should be added to the credits dialog ingame 20081123 17:37:20 so that lines there are wrapped automatically, too 20081123 17:38:07 since the solution of wrapping by hand is, uhm, not nice 20081123 17:38:17 Ivanovic: i agree 20081123 17:38:29 it does not work with different resolution (smaller or bigger) and translations might be *interesting*, too 20081123 17:38:39 Ivanovic: where should people wrapp it by hand? I don't know your screen resolution. 20081123 17:38:40 fendrin_: write a bugreport 20081123 17:39:40 Ivanovic: is this more a bug report or more a feature request? 20081123 17:39:50 bug 20081123 17:40:40 since expected behaviour for text is to wrap at the end of the screen 20081123 17:45:08 ivanovic * r31016 /trunk/ (5 files in 4 dirs): updated Italian translation 20081123 17:45:41 Ivanovic: done 20081123 17:45:52 thanks 20081123 17:54:21 loonybot: { string s("x"); cout << s[s.size()];} 20081123 17:54:23 /usr/lib/gcc/i686-pc-linux-gnu/4.3.2/include/g++-v4/bits/basic_string.h:723: char& string::operator[](size_t): Assertion '__pos < size()' failed. [+ 1 discarded line] 20081123 17:54:44 mordante: ^^ 20081123 17:56:13 I looked it up to be sure it was valid, will check the standard 20081123 18:00:39 -!- EdB [n=EdB@48.153.82-79.rev.gaoland.net] has joined #wesnoth-dev 20081123 18:02:54 loonybot { basic_string s("z"); const char& c = s[1]; } 20081123 18:02:54 error: missing template arguments before 's' 20081123 18:03:04 loonybot { basic_string s("z"); const char& c = s[1]; } 20081123 18:03:05 /usr/lib/gcc/i686-pc-linux-gnu/4.3.2/include/g++-v4/bits/basic_string.h:723: char& string::operator[](size_t): Assertion '__pos < size()' failed. [+ 1 discarded line] 20081123 18:04:54 well seems g++ screws up ;) 21.3.4 "if pos == size(), the const version returns charT()." 20081123 18:06:12 only wonder why it calls the non-const version :/ 20081123 18:06:44 -!- BenUrban_ is now known as BenUrban 20081123 18:08:47 loonybot { const basic_string s("z"); const char& c = s[1]; 20081123 18:08:47 error: Unexpected end of request. 20081123 18:08:49 loonybot { const basic_string s("z"); const char& c = s[1]; } 20081123 18:08:50 20081123 18:09:14 that's one huge const - nonconst difference if I ever saw one 20081123 18:09:26 loonybot { const basic_string s("z"); const char& c = s[1]; cout << c; } 20081123 18:09:27 20081123 18:09:28 well it's allowed 20081123 18:09:45 loonybot { const basic_string s("z"); const char& c = s[1]; cout << hex << c; } 20081123 18:09:46 20081123 18:09:46 the non const version is UB 20081123 18:09:59 loonybot { const basic_string s("z"); const char& c = s[1]; cout << hex << (int)c; } 20081123 18:10:00 0 20081123 18:10:03 mordante: which is quite a major difference 20081123 18:10:24 I agree 20081123 18:10:34 I guess you could use .c_str()[1] 20081123 18:10:56 don't know where and how you're using this anyway ;) 20081123 18:11:24 ilor_: To determine that windows path is absolute :) 20081123 18:11:59 testing for the string length beforehand seems an easy fix 20081123 18:12:09 Santi just commited 3 more scenarios to DM. I guess esr wasn't right about DM being String stable. 20081123 18:12:38 ilor yes of course, just this seemed to be safe as well 20081123 18:13:14 my stl pocket reference doesn't list the subtile difference 20081123 18:13:43 and often it's faster for looking up as the standard ;) 20081123 18:14:11 loonycyborg who's going to fix it? 20081123 18:14:27 would the sorting of the campaigns by the time they play at be a valid feature request? 20081123 18:14:56 atm they're sorted by difficulty 20081123 18:15:14 mordante: that is good as well 20081123 18:15:17 shadowmaster * r31017 /trunk/src/intro.cpp: * Changed Doxygen comments a bit 20081123 18:15:59 fendrin_: make a feature request to allow sorting them by name/difficulty/timeline/dave's preference 20081123 18:16:20 ilor_: lol 20081123 18:16:29 -!- VladimirSlavik [n=chatzill@136.219.broadband6.iol.cz] has left #wesnoth-dev [] 20081123 18:16:49 a small order-by dropdown is always nice when the list of anything is longer than a few items 20081123 18:16:49 fendrin_: well, I'm undecided on the approach: either wrapping in the engine (which makes more sense, considering variable viewport sizes) or adding newlines at arbitrary positions 20081123 18:17:07 mordante: What would be the right way of fixing that? c_str()? 20081123 18:17:34 AI0867: fixing in the engine 20081123 18:17:38 loonycyborg: c_str() iirc is defined to be a null-terminated string so it's okay to take the last element 20081123 18:18:09 c_str() or test for the proper length 20081123 18:18:16 c_str() is null-terminated 20081123 18:18:24 -!- Shadow_Master [n=shadowm@unaffiliated/shadowmaster/x-275924] has joined #wesnoth-dev 20081123 18:18:46 loonybot: { string s("x"); cout << s.c_str()[1];} 20081123 18:18:46 20081123 18:18:46 AI0867: Or you need to add the newlines that often that it fits the lowest possible resolution wesnoth is playable with. 20081123 18:19:34 shadowmaster * r31018 /trunk/ (changelog src/intro.cpp): 20081123 18:19:34 * Removed text wrapping around background image in story screens; it is 20081123 18:19:34 better to always use the whole screen width if possible (fixes the 20081123 18:19:34 first story screens in Descent into Darkness and Under the Burning Suns). 20081123 18:19:34 * Tweaked the algorithm that extends the darkened area below long 20081123 18:19:37 story texts so that it gives a more consistent appearance. 20081123 18:19:46 AI0867: what again could be variable when the font is changed 20081123 18:19:48 zookeeper: I fixed your bug with DiD 20081123 18:20:14 .... damn, the logs today are *long* for a Sunday 20081123 18:20:51 Shadow_Master, awesome 20081123 18:20:59 -!- Sirp [n=user@c-76-102-104-121.hsd1.ca.comcast.net] has joined #wesnoth-dev 20081123 18:21:11 hi Sirp 20081123 18:21:22 -!- Mica [n=Manic_Mi@d14-69-222-189.try.wideopenwest.com] has joined #wesnoth-dev 20081123 18:21:24 hi Ivanovic 20081123 18:21:27 hi Sirp 20081123 18:21:33 Hi Sirp. 20081123 18:21:35 hi Mica 20081123 18:21:41 can anyone tell me how to add a .c file to build with scons? 20081123 18:21:50 Soliton, Ivanovic, anyone: i'd have time for that ANL testing now 20081123 18:21:54 Sirp: add it in SConstruct 20081123 18:22:01 Hi. 20081123 18:22:05 okay 20081123 18:22:12 uhm, src/SConstruct (or something like this) 20081123 18:22:23 zookeeper: i don't know how to use it... 20081123 18:22:23 zookeeper: Great, want some help? I'm maintaining the UMC part of it, if you want help with the mainline part, I'd be glad to. 20081123 18:22:30 -!- crimson_penguin [n=ben@64.201.60.217] has joined #wesnoth-dev 20081123 18:22:33 -!- kthakore2 [n=kthakore@CPE001310a1899c-CM001868e2ad12.cpe.net.cable.rogers.com] has quit ["Leaving"] 20081123 18:22:38 but i am sure Sirp would jump in so that he can see what the problem actually is 20081123 18:22:46 hi crimson_penguin 20081123 18:22:58 hi Ivanovic 20081123 18:22:59 crimson_penguin: which version of gettext utils are you using to create the mo files? 20081123 18:23:07 shadowmaster * r31020 /trunk/src/ (17 files): * Hopefully reduced cohesion a bit by moving the end_*_exception classes/structs out from menu_events.hpp. 20081123 18:23:08 zookeeper * r31019 /trunk/data/multiplayer/scenarios/ANL_utils/ANL_worker_options.cfg: Removed an [allow_undo] which seems to have been forgotten in. 20081123 18:23:22 fendrin_: which gets downright ugly for people with large resolutions, as non-tinygui goes down to 800x480 20081123 18:23:33 Ivanovic: add it where in which SConstruct? :) 20081123 18:23:37 r31020 is your mandatory full-recompile bomb 20081123 18:23:39 AI0867: agreed 20081123 18:23:50 Sirp: ask loonycyborg about the details... 20081123 18:23:52 Ivanovic: msgfmt (GNU gettext-tools) 0.14.5 20081123 18:23:52 we love you too Shadow_Master :P 20081123 18:24:06 crimson_penguin: grml, not good... 20081123 18:24:10 ahhh 20081123 18:24:17 loonycyborg: around so you can tell me? :) 20081123 18:24:18 I'd actually like to see crimson_penguin or other people with non-Linux platforms try to recompile and check that nothing went wrong like the last time we moved headers around... 20081123 18:24:19 Ivanovic: no? :-/ how so? 20081123 18:24:20 crimson_penguin: do you know if there is a way for you to update it to >=0.16? 20081123 18:24:30 Sirp: src/SConscript 20081123 18:24:52 i would like to switch the po files we switch to ones listing the previous string that the string is derived from 20081123 18:25:05 Ivanovic: it isn't in fink 20081123 18:25:10 crimson_penguin: correct 20081123 18:25:14 Ivanovic: this isn't a problem for 1.5.6, is it? 20081123 18:25:29 crimson_penguin: what about this other system to get packages? (don't know the name atm) 20081123 18:25:33 loonycyborg: okay, what do I add in there? 20081123 18:25:36 crimson_penguin: no, for 1.5.6 there is no problem 20081123 18:25:39 MacPorts? 20081123 18:25:43 ok, good :) 20081123 18:25:53 -!- zookeeper2 [n=l@dsl-tkubrasgw1-fe11fa00-51.dhcp.inet.fi] has joined #wesnoth-dev 20081123 18:25:55 or, something like that anyway 20081123 18:26:14 crimson_penguin: right, mac ports 20081123 18:26:15 zookeeper: k, me too. 20081123 18:26:19 oh what i'd give for an apartment-wide UPS. 20081123 18:26:38 anyway, I have it too, but I never really got it to work, and I sort of would rather not anyway, because having two package managers on an OS that isn't really made for package mangers in the first place... just seems a bit messy 20081123 18:26:58 Sirp: A filename to one of lists. 20081123 18:27:00 but maybe I could download the source and install manually 20081123 18:27:30 Soliton, i'll be on the trunk server in a minute.. 20081123 18:27:32 so it's official: my computer IS going to need to be fixed 20081123 18:27:34 crimson_penguin: yes, this should work 20081123 18:27:38 zookeeper: 1.5.6 or trunk? 20081123 18:27:45 obviously it's fine right now, but my graphics card is definitely doomed :( 20081123 18:27:47 1.5.6 should be same? 20081123 18:27:50 crimson_penguin: what is broken in your computer` 20081123 18:27:58 trunk i guess, i don't have 1.5.6 yet. i don't know if the server is the same? 20081123 18:28:11 I think it isn't 20081123 18:28:16 Ivanovic: graphics card; which of course, is directly connected to the motherboard (laptop) 20081123 18:28:30 ugh, this is bad 20081123 18:28:46 nVidia GeForce 8600M GT 20081123 18:29:01 zookeeper2: it's not. ok trunk then. 20081123 18:29:08 yeah; I won't have to pay for it at least, but I'll probably be without a computer for 2 weeks or so :-/ 20081123 18:30:07 in System Profiler, it sometimes lists my PCIe Land Width as x1, sometimes x8 (currently x1) - but it's supposed to be x16 20081123 18:30:38 :( 20081123 18:30:53 crimson_penguin: sounds bad 20081123 18:31:03 Apple has acknowledged the issue: http://support.apple.com/kb/TS2377 20081123 18:31:17 that is at least something 20081123 18:31:19 Ivanovic: 0.16, if I remember correctly. 20081123 18:31:21 and besides that I have AppleCare anyway, so 20081123 18:31:27 so, is MacOS X a cursed platform? we tend to lose their packagers for chaotic reasons such as this 20081123 18:31:43 they never acknowlegdged the problem with ibooks and the power button breaking when using it too often 20081123 18:31:49 Shadow_Master: haha, well, this won't stop me from being a packager - just might make me late once 20081123 18:32:01 Soliton, just a moment, i'll have to revert to some earlier rev first.. 20081123 18:32:04 Ivanovic: oh? that's exactly what happened to my non-Mac laptop 20081123 18:32:09 Ivanovic: well, that would've been their fault - this one they can blame on nVidia ;) 20081123 18:32:12 -!- zookeeper [n=l@dsl-tkubrasgw1-fe11fa00-51.dhcp.inet.fi] has quit [Read error: 145 (Connection timed out)] 20081123 18:32:19 crimson_penguin: ah, right 20081123 18:32:23 I'm currently using it just by touching a part of the circuit 20081123 18:32:59 so, when is 1.5.6 gonna be announced? 20081123 18:33:14 crimson_penguin: some ages ago i wrote a request for fink to update gettext to 0.16* 20081123 18:33:17 http://sourceforge.net/tracker/?func=detail&atid=371315&aid=1741878&group_id=17203 20081123 18:33:19 Sirp: by the way, wanna test my build again, just to be sure? 20081123 18:33:27 but looks like the maintainer is too afraid to do so 20081123 18:33:38 crimson_penguin: and i wanted to announce in some hours 20081123 18:35:05 ok, cool 20081123 18:35:43 * Rhonda has OSX running on her PowerBook ... 20081123 18:35:43 is Darian Lanx really the maintainer? 20081123 18:35:54 Rhonda: what version? :) 20081123 18:36:00 crimson_penguin: 10.5 20081123 18:36:09 Upgraded some months ago. 20081123 18:36:17 Rhonda: ah ha; wanna test my build? 20081123 18:36:41 so far it's been tested on Intel 10.5, and PPC 10.4; so PPC 10.5 would be nice 20081123 18:36:41 Maybe ... at a different time? But yeah, why not. 20081123 18:36:59 Ah, so I take it it is a unversal binary. :) 20081123 18:37:21 Of course :) 20081123 18:37:44 http://downloads.sourceforge.net/wesnoth/Wesnoth_1.5.6.dmg?use_mirror= 20081123 18:37:57 loonycyborg: seems to work, thanks. 20081123 18:38:36 Soliton, still trying... 20081123 18:39:20 -!- ivan_i [n=chatzill@ppp91-76-32-237.pppoe.mtu-net.ru] has joined #wesnoth-dev 20081123 18:39:57 * crimson_penguin thinks SF is ridiculous 20081123 18:40:12 when you first load the releases page, it shows a Downloads column - then after a bit, it disappears 20081123 18:40:36 loonycyborg * r31021 /trunk/src/game.cpp: Corrected a corner case of single-character strings when checking whether path is absolute on windows. 20081123 18:40:39 hrm, seems git-svn is downloading a lot of old revisions because of that annoying utbs branch for some reason.. 20081123 18:40:49 ah, I see; you can turn on/off columns 20081123 18:41:21 Soliton: hm.. It doesn't for me. 20081123 18:41:44 * Espreon laughs manically at Soliton. 20081123 18:41:46 crimson_penguin: Would need to reboot, usually I'm on linux. And should be at a concert in less than two hours ... 20081123 18:42:08 Oh ok - well don't worry about it :) 20081123 18:42:08 Soliton, meh, i'd have to go to a rev older than 1.5.6 to get the latest binary working i guess, so maybe i'll just download 1.5.6 instead 20081123 18:42:20 (unless loony makes a new one right now ;) 20081123 18:42:26 -!- zookeeper2 is now known as zookeeper 20081123 18:42:40 loonycyborg: http://nopaste.com/p/aQknkCwkS 20081123 18:43:07 crimson_penguin: And before I test your package, I most propably should prepare my package (Debian), so that people won't get upset. :P 20081123 18:43:19 I've never got such messages from git-svn 20081123 18:43:27 Rhonda: Ah, good idea :P 20081123 18:43:28 :) 20081123 18:43:32 Or it was so long ago I don't remember.. 20081123 18:43:37 loonycyborg: and then it starts downloading revisions starting from 11000 somewhere. 20081123 18:44:01 * Espreon continues his manical laughter. 20081123 18:44:02 Did I get that right, new campaign in that release? 20081123 18:44:03 Ivanovic: by the way, any approximate time for next 1.4.x? 20081123 18:44:18 Soliton: Does it happen on the server or on your local system? 20081123 18:44:22 might take a while 20081123 18:44:25 loonycyborg: both. 20081123 18:44:25 ;) 20081123 18:44:42 basically i have not planned one anytime soon 20081123 18:44:42 can someone who is competent with art take care about http://www.wesnoth.org/forum/viewtopic.php?f=9&t=22107 20081123 18:44:46 please? 20081123 18:45:13 crimson_penguin: Do you think there is enough potential fixes in 1.4 branch that warrants a release soonish? 20081123 18:45:26 Rhonda: I have absolutely no idea :D 20081123 18:45:36 there are almost no changes at all in 1.4.x 20081123 18:45:55 * Rhonda reboots to OSX anyway. 20081123 18:45:56 Ivanovic: so then, would it be good for me to make a package of the current one? both to get into packaging it, and so people who want one, can use a .app 20081123 18:46:04 Not to test crimson_penguin's package, sorry. ;) 20081123 18:46:14 crimson_penguin: hmm, no idea 20081123 18:46:33 Soliton: I didn't get those messages when compiling win32 binaries. 20081123 18:46:35 crimson_penguin: You mean there is no 1.4.6 for OSX? 20081123 18:46:46 there is 1.4.6 for osx 20081123 18:46:58 it just just "installer based", not "app based" 20081123 18:47:04 loonycyborg: because they have nothing to do with compiling? 20081123 18:47:04 ah 20081123 18:47:09 crimson_penguin: be aware that the package includes python support 20081123 18:47:13 ;) 20081123 18:47:30 Ivanovic: ah yes, that's another challenge 20081123 18:47:36 Soliton: But I always do git svn rebase before compiling :) 20081123 18:47:41 Rhonda: no, there is; it's just an installer, and lots of people don't like installers 20081123 18:48:11 loonycyborg: a rebase does nothing with other branches. 20081123 18:48:33 loonycyborg: fetch is currently still running so don't run it as well. 20081123 18:48:58 I do only rebase. 20081123 18:49:10 What command did you use? 20081123 18:49:29 git-svn fetch 20081123 18:50:01 I never use git-svn fetch so that would explain it :) 20081123 18:50:04 now it starts again with the next stupid let's move utbs around for no reason... 20081123 18:51:08 at least it shouldn't increase repo size since it should all already be there. 20081123 18:51:23 Now wesnothd-1.4 consumes 100% of server's time. 20081123 18:51:50 *CPU time 20081123 18:52:08 * ivan_i uses git svn rebase (and doesn't even track any branches and revisions before 29000) 20081123 18:52:23 Soliton: That branch has a purpose now. 20081123 18:52:27 So NO. 20081123 18:53:03 i guess you didn't read todays discussion. 20081123 18:53:46 Espreon: fendrin agreed to nuke it because esr became angry :) 20081123 18:54:04 Espreon: please read the chat log. The idea of having a branch for UtBS again was heavily bashed by not only esr. 20081123 18:54:20 i'd like to think because it's unnecessary. 20081123 18:54:24 Sorry... 20081123 18:54:38 * Espreon fell asleep on his comp. 20081123 18:54:38 Espreon: not your fault. I should have told you. 20081123 18:55:23 If you want separate branch, better make in wesnoth-umc-dev :) 20081123 18:55:43 Fine. 20081123 18:55:51 I will make one... later. 20081123 18:56:18 So... was that branch nuked or is it going to be nuked soon? 20081123 18:56:34 Shadow_Master: compiling now 20081123 18:57:05 Espreon: it was moved back to trunk and is in 1.5.6 20081123 18:57:13 I know. 20081123 18:57:39 But the branch was updated so fendrin and I could continue to work on the story. 20081123 18:57:45 *from there. 20081123 18:57:53 if we really want a branch were experimental work on mainline can go we can make a real experimental branch but i don't see that as necessary atm. 20081123 18:59:40 -!- Elvish_Pillager [n=eli@24-177-39-196.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev 20081123 19:00:59 -!- Baufo [n=thomas@62-47-138-14.adsl.highway.telekom.at] has joined #wesnoth-dev 20081123 19:02:02 mordante: IMO it would be a good idea to add a function to filesystem.cpp to determine whether path is absolute. 20081123 19:02:23 Shadow_Master: two errors 20081123 19:02:34 src/gui/widgets/debug.cpp:303: error: 'const class gui2::twidget' has no member named 'last_best_size_' 20081123 19:02:49 src/gui/widgets/debug.cpp:342: error: 'const class gui2::tvertical_scrollbar_container_' has no member named 'content_last_best_size_' 20081123 19:04:10 -!- noy [n=Noy@d75-157-52-251.bchsia.telus.net] has joined #wesnoth-dev 20081123 19:04:32 -!- EdB [n=EdB@48.153.82-79.rev.gaoland.net] has quit ["Konversation terminated!"] 20081123 19:05:46 crimson_penguin don 20081123 19:05:54 't compile that file 20081123 19:06:02 ah, ok :) 20081123 19:06:06 thought that might be the case 20081123 19:06:32 maybe I should add guards around it 20081123 19:06:57 those 'missing' members are also in #ifdefs 20081123 19:07:03 no problems then :) 20081123 19:07:18 loonycyborg might be an option 20081123 19:07:31 not sure where else we use it 20081123 19:13:25 -!- zookeeper2 [n=l@dsl-tkubrasgw1-fe11fa00-51.dhcp.inet.fi] has joined #wesnoth-dev 20081123 19:14:13 gah, power shortages. 20081123 19:14:32 Soliton, i installed 1.5.6 so we can use that now 20081123 19:15:08 but please comment out the [allow_undo] in ANL_worker_options.cfg:51 first 20081123 19:16:06 -!- Elvish_Pillage2 [n=eli@24-177-39-196.dhcp.oxfr.ma.charter.com] has quit [Read error: 110 (Connection timed out)] 20081123 19:16:54 -!- zookeeper [n=l@dsl-tkubrasgw1-fe11fa00-51.dhcp.inet.fi] has quit [Read error: 145 (Connection timed out)] 20081123 19:17:03 uh, that and also ANL_research_options.cfg:191 and ANL_help.cfg:7 20081123 19:17:10 weird, i thought i removed those a long time ago. 20081123 19:17:38 -!- zookeeper2 is now known as zookeeper 20081123 19:18:02 done. 20081123 19:20:06 mordante * r31022 /trunk/src/game.cpp: Also use the c_str() at the relative path. 20081123 19:25:26 mordante * r31023 /trunk/src/gui/widgets/ (scroll_label.cpp scroll_label.hpp): 20081123 19:25:26 Fix the scrolling bar for the scroll_label. 20081123 19:25:26 It was broken in the layout engine rewrite. Testing didn't reveal the problem 20081123 19:25:26 since it's not used yet. 20081123 19:32:33 -!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev 20081123 19:37:49 -!- Mica [n=Manic_Mi@d14-69-222-189.try.wideopenwest.com] has quit ["What did you say? I can't hear you, my connection's gone."] 20081123 19:40:21 -!- crimson_penguin [n=ben@64.201.60.217] has quit ["Leaving"] 20081123 19:55:03 bye 20081123 19:55:12 -!- fendrin_ [n=fabi@g228067232.adsl.alicedsl.de] has quit [Remote closed the connection] 20081123 20:02:35 mordante * r31024 /trunk/src/gui/widgets/ (10 files): Make the text wrapping work. 20081123 20:11:10 -!- zookeeper [n=l@dsl-tkubrasgw1-fe11fa00-51.dhcp.inet.fi] has quit [Read error: 145 (Connection timed out)] 20081123 20:11:15 -!- Espreon [n=espreon@c-76-117-239-33.hsd1.pa.comcast.net] has quit [Remote closed the connection] 20081123 20:11:47 -!- wesbot changed the topic of #wesnoth-dev to: released 1.5.6, announcing soon... | soft stringfreeze for most campaigns: http://www.wesnoth.org/wiki/TextdomainStatus | List 1.6 features here: http://www.wesnoth.org/wiki/Wesnoth1.6Features | 50 bugs, 209 feature requests, 9 patches | logs: http://wesnoth.org/irclogs | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com, or for images: http://imagebin.org/index.php?page=add 20081123 20:11:55 -!- me [n=me@athedsl-4507893.home.otenet.gr] has joined #wesnoth-dev 20081123 20:12:01 -!- me is now known as santi 20081123 20:12:32 -!- santi is now known as kalenzz 20081123 20:14:36 -!- IRSeekBot3 [n=IRSeekLo@irseek/log-bot/IRSeekBot] has quit [K-lined] 20081123 20:17:41 -!- Baufo [n=thomas@62-47-138-14.adsl.highway.telekom.at] has quit ["Leaving"] 20081123 20:19:53 -!- zookeeper [n=l@dsl-tkubrasgw1-fe11fa00-51.dhcp.inet.fi] has joined #wesnoth-dev 20081123 20:20:39 am I reading wrong, or someone wants to replace me with a web interface for translations? :) 20081123 20:21:37 crimson_penguin: roger 20081123 20:21:55 oh wait, hat's Mordante's fault ;) 20081123 20:26:42 kalenzz: hi... 20081123 20:27:40 well, looks like ANL works fine now. 20081123 20:27:54 we played a couple of turns and got no OOS. 20081123 20:28:18 zookeeper: maybe those [allow_undo]s messed it up before? 20081123 20:29:22 Soliton, could be, although i'm pretty sure that last time i tried to make sure they couldn't be a problem 20081123 20:29:47 hi Shadow Master 20081123 20:30:33 -!- thespaceinvader [n=chatzill@91.108.187.175] has joined #wesnoth-dev 20081123 20:31:07 zookeeper: yeah, i actually think so too. 20081123 20:54:23 shadowmaster * r31025 /trunk/data/core/about.cfg: * Remove reference to inexistent, typoed image file 20081123 20:55:23 ee 20081123 20:55:29 oops, wrong tab 20081123 20:56:09 wow, svn 1.5 is impressive 20081123 20:56:36 Shadow_Master: in what aspect ? 20081123 20:57:10 it asks me what to do with individual conflicts wihile svn up'ing 20081123 20:58:06 should we remove the "campaign design" entry entirely from the core credits? 20081123 20:58:16 and has a good deal of options on what-to-do at that point 20081123 20:58:28 writing the 1.5.6 announcement now 20081123 20:58:30 * loonycyborg aswers postpone to all such svn questions. 20081123 20:58:42 zookeeper: I think so, and merge missing entries onto each campaign's credits 20081123 20:58:49 Shadow_Master: yes, the latest version of subversion is *great* 20081123 20:58:58 really lovely when you got a conflict at svn up 20081123 20:59:39 s/aswers/answers 20081123 20:59:42 that is, I think I'veseen one or two entries there not on the respective campaign's credits 20081123 21:07:05 -!- crimson_penguin [n=ben@64.201.60.217] has joined #wesnoth-dev 20081123 21:08:12 -!- PK [n=pk@r74-192-44-206.vctrcmta01.vctatx.tl.dh.suddenlink.net] has joined #wesnoth-dev 20081123 21:08:34 yo kittens 20081123 21:09:10 wesbot: seen kitty 20081123 21:09:11 boucman: Sorry, I don't know of kitty. 20081123 21:09:15 come here with daddy ^__^ 20081123 21:09:26 Shadow_Master: at least one kitten's not here :) 20081123 21:11:22 -!- shikadibot [n=ai@van-gessel.demon.nl] has joined #wesnoth-dev 20081123 21:11:27 -!- zookeeper [n=l@dsl-tkubrasgw1-fe11fa00-51.dhcp.inet.fi] has quit [] 20081123 21:11:33 * PK hopes he's not the subject of conversation 20081123 21:11:43 shikadibot: PK? 20081123 21:11:44 Sorry, I don't know what 'PK' means... 20081123 21:11:54 shikadibot: learn: "PK PsychoticKittens" 20081123 21:12:02 shikadibot: winr? 20081123 21:12:02 WINR: Wesnoth Is Not Realistic 20081123 21:12:05 :) 20081123 21:12:19 shikadibot: riplib? 20081123 21:12:19 RIPLIB: Reduction In Power when Leveling Is Bad 20081123 21:12:25 shikadibot: part 20081123 21:12:25 -!- shikadibot [n=ai@van-gessel.demon.nl] has left #wesnoth-dev [] 20081123 21:13:02 fancy doombot, Shadow_Master 20081123 21:13:18 avoid letting it destroy the universe. 20081123 21:13:32 because then I'd be unable to play wesnoth, and that would be bad. 20081123 21:13:56 ... really? 20081123 21:14:18 PK: If you want to have it back, join #wesnoth-umc-dev, nick to ShikadiLord and say shikadibot: join #wesnoth-dev :) 20081123 21:15:16 -!- ShikadiLord [n=shadowm@unaffiliated/shadowmaster/x-275924] has joined #wesnoth-dev 20081123 21:16:07 Soliton: the server for 1.5.6 is up and running? 20081123 21:16:13 now PK is most indefinatly lost. 20081123 21:16:19 all major binaries are done, so the 1.5.5 server could be shut down 20081123 21:17:23 -!- ShikadiLord [n=shadowm@unaffiliated/shadowmaster/x-275924] has quit [Client Quit] 20081123 21:18:39 loonycyborg: there's an easier way to achieve that, and is just to join #wesnoth-umc-dev with one's real nickname 20081123 21:19:38 Shadow_Master: But it's not nearly as fun :) 20081123 21:22:04 1.5.6 announcement: http://www.wesnoth.org/forum/viewtopic.php?f=5&t=23050 20081123 21:23:31 can I proofread it? 20081123 21:24:49 of course! 20081123 21:25:10 is expected to be as smart again as it used to be <- the again looks out of place 20081123 21:25:43 "Another important not is that 'Under the...'" <- should be note 20081123 21:25:55 ivanovic * r31026 /trunk/RELEASE_NOTES: clean up release notes... 20081123 21:26:23 -!- Patterner [n=Psyche@e177114027.adsl.alicedsl.de] has quit [Read error: 145 (Connection timed out)] 20081123 21:26:45 fixed 20081123 21:27:09 otherwise, fine 20081123 21:27:22 (to my standards, that is) 20081123 21:28:45 ... should this be the perfect time for adding the direct download links to the frontpage? 20081123 21:28:55 a.k.a. give 1.5.x more epxposure 20081123 21:30:07 hey, this is what i am working on atm 20081123 21:30:15 give me some time to update the frontpage with the announcement... 20081123 21:30:19 ah :) 20081123 21:32:13 Ivanovic: yes. 20081123 21:34:57 Soliton: great 20081123 21:36:03 Ivanovic: need some help? 20081123 21:36:05 "probably the the AI" 20081123 21:36:56 -!- Patterner [n=Psyche@e177224078.adsl.alicedsl.de] has joined #wesnoth-dev 20081123 21:37:00 "Under the burning Suns" - should be capital B 20081123 21:39:32 crimson_penguin: both fixed 20081123 21:39:37 updated the frontpage 20081123 21:39:46 :) 20081123 21:40:39 everyone: please proofread the frontpage (http://www.wesnoth.org/) and the forum post about the release (http://www.wesnoth.org/forum/viewtopic.php?p=322972#p322972) 20081123 21:41:28 -!- Ivanovic changed the topic of #wesnoth-dev to: soft stringfreeze for most campaigns: http://www.wesnoth.org/wiki/TextdomainStatus | 1.6 release plan: http://www.wesnoth.org/wiki/Wesnoth1.6ReleasePlan | 50 bugs, 209 feature requests, 9 patches | logs: http://wesnoth.org/irclogs | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com, or for images: http://imagebin.org/index.php?page=add 20081123 21:44:30 the frontpage doesn't update/redownload automaticlly :/ 20081123 21:44:38 a PHP issue? 20081123 21:44:52 Ivanovic: very nitpicky, but "Mac OSX" should probably have a space between OS and X 20081123 21:44:58 Shadow_Master: I found that too 20081123 21:45:28 particular of that page, others get redownloaded automatically on changes as expected 20081123 21:45:54 crimson_penguin: made it 'MacOSX' all over the place now 20081123 21:45:59 (front page) 20081123 21:46:18 no spaces? 20081123 21:46:20 'MacOS X' 20081123 21:46:30 crimson_penguin: jupp 20081123 21:46:39 * crimson_penguin thought it should be "Mac OS X" 20081123 21:46:47 takes to much room... 20081123 21:47:06 Ivanovic: 'quite many' shouldn't it be 'quite a lot' ? 20081123 21:47:13 bahaha, ok 20081123 21:47:33 at least it sounds awkward 20081123 21:47:39 * quite a lot of 20081123 21:55:51 okay, afk now that my work for today is done 20081123 21:55:54 ;) 20081123 21:55:55 Ivanovic "Now is the perfect time for translators to start working on the development branch, too." is the comma needed 20081123 21:56:11 mordante: i think so 20081123 21:56:24 180.7 MB vs 147.9. that's really impressive. that's really good exposure if I understand correctly the way the minds of Internet users work nowadays 20081123 21:56:31 -!- VladimirSlavik [n=chatzill@136.219.broadband6.iol.cz] has joined #wesnoth-dev 20081123 21:56:35 also not sure, still reading 20081123 21:59:47 seems you love to meantion 1.5.6 might be more buggy as 1.4.x ;) no other problems found 20081123 22:00:58 -!- Aguy [n=chatzill@cpe-98-14-10-36.nyc.res.rr.com] has joined #wesnoth-dev 20081123 22:00:59 hey, it might be 20081123 22:01:24 I didn't deny that ;) 20081123 22:02:09 I'll try to keep an eye on Jetryl's commiits to have a complete list of graphic changes from now on 20081123 22:03:02 that would be nice :) 20081123 22:03:05 mordante: your widgets aren't useful yet for replacing the story screen code, right? 20081123 22:03:38 not yet, I just finished the wrapping code tonight, which is a back step forward 20081123 22:03:50 it is currently a bit obscure and obfuscated 20081123 22:04:39 and I'm brainstorming on how to make it look a bit more consistent with the rest of Wesnoth's interface 20081123 22:04:40 I need to add a way to make it "typing" the text 20081123 22:05:14 i.e. adding those nifty line borders on the text overlays for title and/or lower text 20081123 22:05:15 mordante: displaying all of it would be a good start 20081123 22:05:23 that is: currently it is ugly and broken... 20081123 22:05:24 ;) 20081123 22:06:02 I first want to get the dialogs done, then I at least have the feeling something useful has been done ;) 20081123 22:07:25 also I want to focus on my 1.6 todo list first ;) 20081123 22:08:13 but the story screen seems a recent regression so it would be a good idea to see what caused it 20081123 22:09:29 due to this regession it might make sense to move it up in the list of targets 20081123 22:09:44 though maybe with just a plain "all text appears at once" approach 20081123 22:10:27 the dialogs are first, the portraits look to awesome, not to use them ;) 20081123 22:10:39 :) 20081123 22:10:56 mordante: what occurred to the story screen? 20081123 22:11:09 also depending on Kitty's verdict I might need to do more WML changes 20081123 22:11:40 they sometimes get an ugly coloured pattern, there's something in yesterdays log 20081123 22:11:41 -!- kalenzz is now known as santi 20081123 22:12:18 -!- zookeeper [n=l@dsl-tkubrasgw1-fe11fa00-51.dhcp.inet.fi] has joined #wesnoth-dev 20081123 22:12:24 and I think it might have something to do with the recent scaling changes in that area, but haven't investigated 20081123 22:13:02 but maybe I'm able to look into the images before 1.6 but no guarantees 20081123 22:13:20 first the dialogs and the gold carry over system, they are important 20081123 22:14:18 wha's the best way to reproduce that behavior? 20081123 22:15:22 -!- Aguy [n=chatzill@cpe-98-14-10-36.nyc.res.rr.com] has quit [Read error: 104 (Connection reset by peer)] 20081123 22:15:50 -!- IRSeekBot3 [n=IRSeekLo@ip-68-178-158-148.ip.secureserver.net] has joined #wesnoth-dev 20081123 22:16:55 -!- santi is now known as fnaek 20081123 22:17:37 Shadow_Master it depends on your screen size, now using 800x600 and start UtBS 20081123 22:19:14 I can't see anything weird 20081123 22:19:40 remember that I commited a fix for the text wrapping in sstory screens earlier today 20081123 22:19:42 you started wesnoth in windowed mode 800x600? 20081123 22:20:27 (tested UtBS at HEAD, 800x600 fullscreen) 20081123 22:20:46 tested with r31024 20081123 22:21:09 no windowed not fullscreen 20081123 22:22:31 testing something that xwininfo claims to be 800x600 20081123 22:22:52 nothing weird besides the broken area of the LCD 20081123 22:22:58 I used wesnoth -r 800x600 20081123 22:23:51 still looks fine to me (r31025) 20081123 22:23:55 (O0) 20081123 22:24:17 I'll put an O3 build to make in the background 20081123 22:25:08 also O0 odd 20081123 22:26:19 Using debian lenny 20081123 22:26:36 that is, libpng 1.2.27 20081123 22:26:55 also on lenny 20081123 22:27:08 are you perhaps using some weird X server layout? (compiz? or some of the other fanciness addons) 20081123 22:27:31 a screenshot would be appreciated too 20081123 22:27:36 no as plain as it gets using fvwm nothing fancy 20081123 22:28:27 I've a hunch, 1 sec 20081123 22:30:03 interesting, changed from German to English and now it looks different (still bad) screeny will be up soon 20081123 22:30:53 to difference is that it went from black to blue (you'll understand when you see the screeny) 20081123 22:31:06 s/to/the/ 20081123 22:31:09 * boucman raises an eyebrow 20081123 22:31:24 how could language cause graphical glitches with images ? 20081123 22:31:56 I assume it's size related and German and English have different text lengths 20081123 22:34:15 an O3 build with more intrusive optimizations still works fine 20081123 22:34:54 here she is http://imagebin.org/31995 20081123 22:36:22 I'd say "memory corruption" 20081123 22:36:41 I'd place my suspicions on the WM 20081123 22:36:52 yes that also came to my mind, want to test with valgrind 20081123 22:37:02 if you can reproduce easily, bisecting is probably the way to go 20081123 22:37:11 but that won't be tonight since it takes too long 20081123 22:37:14 or, the display device 20081123 22:37:23 I'm not sure how reproducable it is 20081123 22:37:35 I think I once experienced similar stuff when having a driver configured to use EXA instead of XAA 20081123 22:37:37 we had several reports of it 20081123 22:37:43 oh 20081123 22:44:43 I'm off now, night 20081123 22:44:55 -!- mordante [n=chatzill@roadie.xs4all.nl] has quit ["ChatZilla 0.9.84 [Firefox 3.0.3/2008092417]"] 20081123 22:48:22 -!- noy [n=Noy@d75-157-52-251.bchsia.telus.net] has quit [] 20081123 22:48:53 -!- ivan_i [n=chatzill@ppp91-76-32-237.pppoe.mtu-net.ru] has quit ["ChatZilla 0.9.84 [Firefox 3.0.4/2008111317]"] 20081123 22:51:30 -!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit [Remote closed the connection] 20081123 22:53:45 -!- crimson_penguin [n=ben@64.201.60.217] has quit ["Leaving"] 20081123 22:57:14 -!- fendrin [n=fabi@g228067232.adsl.alicedsl.de] has joined #wesnoth-dev 20081123 22:58:24 hi fendrin 20081123 22:58:31 this is santi 20081123 22:58:57 hi fnaek 20081123 22:59:05 how are you? 20081123 22:59:40 nice work on DM it looks promising 20081123 22:59:56 I was lookingat balancing issues in LoW-maybe in scenario 2 the trolls should target elves more heavily than dwarves on hard 20081123 23:00:37 DM is finished scenariowise, but it needs testing and fixing for WML 20081123 23:01:51 -!- zookeeper [n=l@dsl-tkubrasgw1-fe11fa00-51.dhcp.inet.fi] has quit [] 20081123 23:01:51 fnaek: I already started fixing wml things in DM 20081123 23:02:01 fnaek: I have an issue with Sagas's name 20081123 23:02:06 scenario 12 is already rewritten to some degree 20081123 23:02:30 AI0867 you mean Sagus? 20081123 23:02:31 sagus* 20081123 23:02:35 yes 20081123 23:02:41 ok, what's the issue? 20081123 23:02:43 not the Sagus part, but the Mal part 20081123 23:02:56 Mal-Ravanal didn't exist yet 20081123 23:02:58 so no Mal 20081123 23:03:04 You are fast! You already got to that part 20081123 23:03:16 Ok, Sagus-Malal? 20081123 23:03:29 fnaek: I read the commits 20081123 23:03:35 which means I see changes immediately 20081123 23:03:54 Yes, but you also need to at least have read the cfg 20081123 23:04:13 true, and I did 20081123 23:04:13 diff. 20081123 23:04:22 I think our ML has diffs 20081123 23:04:37 umc-dev doesn't 20081123 23:04:40 wesnoth does 20081123 23:04:45 though I can easily switch us over 20081123 23:04:54 or... I think I disabled them actually, for being often too large and forcing me to go to approve the mails 20081123 23:07:08 ilor_: the forest icon in the editor extends off the side of the hex if in the left-mouse-button location 20081123 23:07:20 meaning it leaves a trail as you move the mouse around 20081123 23:07:27 Shadow_Master: the gna lists somehow automatically reduce to "no diff" when the diff would get too big 20081123 23:07:29 fnaek: about the LoW scenario2. The change should be easy. I can commit it within an hour but I can't playtest it in the near future. 20081123 23:07:52 AI0867: parsing what you just said 20081123 23:08:02 Ivanovic: not a feature in SF.net's ML software (mailman) IIRC 20081123 23:08:08 :( 20081123 23:08:24 er. the svn hook 20081123 23:08:24 fendrin, ok, I will playtest; it won't be backported to the 1.5.6 version that's out 20081123 23:08:52 fnaek: no that can't be done. 20081123 23:08:59 AI0867: ok can reproduce, will have to crop the overlay to the hex size 20081123 23:09:33 fendrin, I know. Any changes will be in future releases 20081123 23:10:56 fnaek: And I still haven't put a cool musik selection to some of the scenarios. 20081123 23:11:18 I guess that work will be done until 1.5.7 20081123 23:11:23 ilor_: you didn't really fix the terrain config duplication in editor2 20081123 23:11:35 Shadow_Master: I didn;t? 20081123 23:11:50 ilor_: it mutates into a more visible problem if you enter the editor (with add-ons capable of editor code) and then go back to the campaigns menu 20081123 23:11:56 Fendrin, you mean in Low or DM? 20081123 23:12:04 low 20081123 23:12:12 add-on entries get duplicated in the campaigns menu then 20081123 23:12:43 ok, which scenaria? 20081123 23:12:54 ok that's really weird 20081123 23:13:00 I think something is trying to parse the campaigns config twice without clearing the config object, whiochever it is 20081123 23:13:18 do you mind if I take a look at it and touch your code ? :) 20081123 23:13:27 Shadow_Master: go ahead 20081123 23:13:39 Shadow_Master: at least the editor group itself is not duplicated ;) 20081123 23:13:56 yeah, but the terrains still appear twice in the aplette 20081123 23:14:13 at least in the "All" group 20081123 23:14:20 fnaek: shall I post a list with the currently used music tracks in the forum? 20081123 23:14:50 -!- miq [n=miq@krlh-5d83b435.pool.einsundeins.de] has quit [Remote closed the connection] 20081123 23:15:13 there's more than one problem involved here 20081123 23:15:20 fendrin, why not? 20081123 23:15:28 1) the editor should ignore duplicate terrain types I think 20081123 23:16:04 fnaek: okay, give me a minute 20081123 23:16:07 2) it seems the editor includes the addons twice, look for the ifdefs along the editor path in core/_main.cfg maybe theres something which shouldn't be there 20081123 23:16:21 3) things from the editor pollute the main menu data 20081123 23:16:36 -!- boucman [n=rosen@159.83.102-84.rev.gaoland.net] has quit [Remote closed the connection] 20081123 23:16:46 3? what is the symptom? 20081123 23:16:59 oh wait, nvm, I figured it out 20081123 23:17:34 time for me to head off to bed, n8 20081123 23:17:41 good night 20081123 23:18:24 3 I would have epxected to see it fixed by suokko's changes to WML config amangement 20081123 23:20:37 you mean by the changes introducing this bug? https://gna.org/bugs/?12568 20081123 23:21:21 let me check 20081123 23:22:13 nevertheless, 3 shouldn't have been caused by his changes either, I recall this issue with the editor ocurring before that 20081123 23:22:19 ç 20081123 23:22:52 ç? 20081123 23:23:03 Is this some greek letter? 20081123 23:23:51 fnaek: done 20081123 23:25:18 ilor_: the "terrain picker" keeps shrinking for some reason, I now only have 4 icons there, though there's room for plenty more 20081123 23:25:35 loonycyborg: c'est français :) 20081123 23:25:57 my spanish keyboard includes french keys 20081123 23:26:00 -!- VladimirSlavik [n=chatzill@136.219.broadband6.iol.cz] has quit ["ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]"] 20081123 23:26:01 AI0867: the terrain picker gets wacky if you change resolutions around 20081123 23:26:18 -!- VladimirSlavik [n=chatzill@136.219.broadband6.iol.cz] has joined #wesnoth-dev 20081123 23:26:25 it's crap code that needs a rewrite that waits for new widgets 20081123 23:26:34 fendrin -yes, but this is the complete list, not all music tracks have problems, right? 20081123 23:27:06 yes, I just got 4 back by dragging the borders around 20081123 23:28:58 fnaek: "Sagus-Malal" would imply he has something to do with Ihiah-Malal, there is no generic "I have become a lich" pre/postfix 20081123 23:29:18 fnaek: right, I believe some of the levels have a good selection of music. 20081123 23:30:11 AI0867, In HttT, around the same time, you have a bunch of Malals. In DM you have Iliah-Malal, in HttT Muff-Malal and more I do not recall right now 20081123 23:31:35 fnaek: that's a bug, not a standard =P 20081123 23:31:51 it's just one with a low priority to fix 20081123 23:32:49 fnaek: btw, in DM there was a mix of iHiah and iLiah, esr decided to standardise on Ihiah-Malal, but which do you say is the correct one? 20081123 23:34:24 ilor_: the main issue is that it occasionally makes terrains unavailable except through the 'all' list 20081123 23:35:26 AI0867, It should be Iliah-Malal(this was not my design, I am merely respecting the name given by the original design author) 20081123 23:35:30 fnaek: and actually, they're all named 'Muff something' 20081123 23:35:41 one of them just happens to be called Muff Malal 20081123 23:35:54 fnaek: k 20081123 23:36:19 also, I have no strong opinions on this. No problem changing it to whatever. 20081123 23:36:24 want me to change it around? 20081123 23:36:32 or we can just leave it as it is 20081123 23:36:37 Sure, any ideas? 20081123 23:37:49 I'm a bit lost, which one were we talking about again? 20081123 23:38:06 The name of Sagus as a lich, no? 20081123 23:38:27 oh, right 20081123 23:38:38 well, we can keep him as Sagus 20081123 23:38:49 or he could have changed his name to sound more evil 20081123 23:38:55 is this valid wml code? [ai] [target] value=1.5 race=elf 20081123 23:38:57 maybe better to do that 20081123 23:39:08 fendrin where is that? 20081123 23:39:32 better to do that=yes, change Sagus to make it more evil 20081123 23:39:39 fendrin: I believe it is 20081123 23:39:44 fnaek: on my hard disk. If it fits I am going to commit it to scenario2 20081123 23:40:27 ok, I thought it was my new code in DM 20081123 23:40:43 fnaek: so I change Ihiah back to Iliah? 20081123 23:41:38 fendrin * r31027 /trunk/data/campaigns/Legend_of_Wesmere/scenarios/02_Hostile_mountains.cfg: LoW 2: tuned the trolls ai to target the elves stronger than the dwarves. 20081123 23:42:56 AI0867, yes, if you wish. Like I said, unless there is a strong reason, I prefer to respect the name of the original author 20081123 23:43:47 k 20081123 23:44:32 ok, thanks fendrin. I'll playtest tomorrow. Now got to finish some code 20081123 23:45:56 fnaek: another thing: I fixed the map portal.map 20081123 23:46:33 but i had to remove 2 fields which are the borders of that portal. What kind of graphics or terrain should be the portal? 20081123 23:50:27 Maybe something magical, like the utbs stuff. Don't remember what I used originally 20081123 23:50:55 Or maybe an impassable wall. 20081123 23:51:43 Incidentally, on DM, esr(I guess) must have changed Save the King to Save the Prince; this creates update problems, 20081123 23:52:14 i.e. the name of the file must be the same as referenced by the previous scenario and the same as listed on that scenario 20081123 23:52:48 also, it kind of changes thestory, because in scenario 5, it's the King who leaves 20081123 23:52:59 we'll discuss this with esr 20081123 23:53:33 fnaek: if you look at the history of wesnoth, the king is very old at that time 20081123 23:53:44 it's unlikely for him to stand in a battle 20081123 23:56:32 ok, but that was not the original design and probably we need to change scenario 5 too 20081123 23:59:14 fnaek: your r2666 broke the showdown map 20081123 23:59:23 it left svn conflict markers in there 20081123 23:59:59 I'll try to merge fendrin's changes with yours --- Log closed Mon Nov 24 00:00:08 2008