--- Log opened Thu Oct 03 00:00:40 2013
20131003 00:03:14< irker394> wesnoth: mattsc wesnoth-old:master ce958fec50c5 / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: Update Xcode project file http://git.io/jTqc_w
20131003 00:03:22< shadowm> underlying_id is pretty much the poor man's UID solution. No idea if anyone even considered anything more sophisticated than that.
20131003 00:04:18-!- trademark_ [~trademark@nsg93-8-88-175-59-164.fbx.proxad.net] has quit [Ping timeout: 264 seconds]
20131003 00:07:56< shadowm> underlying_id is not guaranteed to be meaningful or constant for a unit subject to serialization, recall carry-over, etc., so WML and Lua coders must use the plain-text id instead for reliable identification of special units.
20131003 00:10:04< gfgtdf> shadowm: but if teh normal "id" is identfyinf why do we need undderlayug id?
20131003 00:10:23< gfgtdf> the*  identyfing*
20131003 00:11:20< justinzane> So, shadowm, since boost is already a dep; and, since boost::uuids is part of the standard boost install (I think.); would it be reasonable for me to try to refactor those so that ...
20131003 00:11:28-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20131003 00:12:17< justinzane> echo unit had a boost::uuids::uuid id_ that is highly likely to be unique for all players even on a rather large server?
20131003 00:12:24< shadowm> gfgtdf: It used to be that the plain id wasn't checked for collisions.
20131003 00:12:56< justinzane> it gets rid of some unexpected code paths
20131003 00:13:47< shadowm> justinzane: I can't really answer that, sorry. The underlying_id mechanism was introduced too long ago and I don't remember the reasoning that led up to it.
20131003 00:14:31< shadowm> Also, note that the large server argument is irrelevant. All units live in their own separate gamestates, one for each MP match.
20131003 00:17:11< gfgtdf> shadowm: now it is. But if someone want for example wants to make a tournamtent with recall list from last scenario or something similar, uniqueness might be useful.
20131003 00:17:30< gfgtdf> s/example/example  if
20131003 00:18:00< shadowm> Hence I mentioned underlying_id is not guaranteed to be unique during the lifetime of a unit.
20131003 00:18:17< shadowm> I mean constant.
20131003 00:18:58< gfgtdf> shadowm: is there a specail reason for that ?
20131003 00:19:02< gfgtdf> scpecial*
20131003 00:19:02< shadowm> What _is_ guaranteed to be constant is the unit's plain id. When left unspecified by the scenario designer, the game automatically generates a plain id like "Unit Type-numbersnumbersnumbers".
20131003 00:20:13< shadowm> It seems that numbersnumbersnumbers right now is the underyling_id as long as it doesn't change.
20131003 00:20:52< shadowm> So when can underlying_id really change? I don't know, those are implementation details I don't handle. I'm mostly just a user when it comes to units. :p
20131003 00:21:09< shadowm> And I've been told to not look at underlying_id.
20131003 00:21:24< gfgtdf> shadowm: by who ?
20131003 00:21:33< shadowm> Why the people who implemented it.
20131003 00:21:40< gfgtdf> hm ok
20131003 00:22:17< shadowm> Because since I'm a user, I don't have any business to do with an ugly implementation detail that is just unfortunately exposed through the WML interface for reasons I don't remember either.
20131003 00:23:33< shadowm> Tracking back the first commit introducing underlying_id and finding the IRC logs and ML posts of around that time would probably help if you really need to know about the design rationale.
20131003 00:23:47< gfgtdf> shadowm: do you know about the code that reads files ?
20131003 00:24:02< shadowm> But again, the unit management code has changed a few times since then (including the addition of the plain id uniqueness requirement).
20131003 00:24:46< shadowm> gfgtdf: Depends. Maybe try with a more specific question?
20131003 00:25:26< shadowm> We currently read WML files through the parser, and WML files through the preprocessor, and Lua files.
20131003 00:26:22< gfgtdf> teh game cache produces corruptes 2-bytes long .gz files, windows crashes when reading them, but other version not do you know why not 
20131003 00:26:45< gfgtdf> corrupted*, s/not/not?
20131003 00:26:51< shadowm> I believe we just pass them to the Boost iostreams pipeline and expect that to work.
20131003 00:27:31< gfgtdf> on visual studio boots thorws gzip_error s and crashes because they were not catched
20131003 00:27:31< shadowm> IIRC there are Boost versions for which that goes wrong in different ways. Is this related to your post in the Technical Support section?
20131003 00:28:22< gfgtdf> the post today ? noi think that was a problem with microsoft windows  itself.
20131003 00:28:37< shadowm> Also, I thought someone had solved the cache corruption issue before.
20131003 00:29:14< gfgtdf> i have made a pull request with different msvc issues, containg a try to fix that.
20131003 00:29:22< shadowm> 535d75d88d376045952d0490bb7df268394fe495 in 1.10, for example.
20131003 00:29:39< shadowm> wesbot: bug 16544
20131003 00:29:40< wesbot> shadowm: Are you sure with that number?  Please check again...
20131003 00:30:26< shadowm> Rhonda: Why is wesbot not reporting bug counts anymore and failing to report bugs that actually exist like https://gna.org/bugs/?16544 ?
20131003 00:30:55< gfgtdf> filter.peek(), sets badbit to true, but not returning EOF.
20131003 00:31:55< gfgtdf> save_index_class::data expects the gzip_error to be thown.
20131003 00:39:49-!- H-Hour [~H-Hour@cpc7-sgyl35-2-0-cust428.18-2.cable.virginmedia.com] has quit [Quit: Leaving]
20131003 00:39:59-!- anonymissimus [~chatzilla@HSI-KBW-149-172-228-192.hsi13.kabel-badenwuerttemberg.de] has joined #wesnoth-dev
20131003 00:40:38< gfgtdf> shadowm: those 2 byte gz files arent empty files they are really corrupt files, other zip progrramms liek 7 also give me an error when opening the,
20131003 00:41:12-!- Turuk [~Turuk@cpe-65-29-217-233.cinci.res.rr.com] has joined #wesnoth-dev
20131003 00:43:07< anonymissimus> gfgtdf: I don't think catching that error is enough, really; I had some session with mordante about that, he has a waiting patch afaik; last status is "please commit it! - answer: "it needs more testing""
20131003 00:43:47< anonymissimus> the bad thing about the error is that MSVC cannot use the wml cache at all, making debugger restarts much slower
20131003 00:47:07< gfgtdf> wesbot: seen mordante
20131003 00:47:08< wesbot> gfgtdf: The person with the nick mordante last spoke 3d 2h ago.  3d 2h ago was here and on the channel #wesnoth-de with the message: Quit: Leaving
20131003 00:48:43< anonymissimus> I also get the save index crash (I think)
20131003 00:48:46< gfgtdf> anonymissimus: well the exeption is thrown so catiching is seems the normal solution, and i dont see any drawbacks of catching ot
20131003 00:49:41< gfgtdf> i think the reason for the save index crash is, that save_index_class::data expects the gzip_error to be thown but read_gz doenst do that
20131003 00:49:52< justinzane> gfgtdf: the use case about "keeping" a unit is something I have wanted as a player, though I think that it is probably inappropriate for the game. After all, when I was a kid I gave up D&D because everyone seemed to have a level 9 mage/fighter/whatever. 
20131003 00:50:23< anonymissimus> MSC_VER is 1500 for me
20131003 00:50:39< gfgtdf> justinzane: what is D&D
20131003 00:50:42< gfgtdf> ?
20131003 00:51:32< gfgtdf> for me it is 1600 i want sure about older version so i made the #if as restrictive as possible
20131003 00:51:38< gfgtdf> wasn't
20131003 00:53:06< shadowm> justinzane: It is perfectly possible atm to implement an add-on to store units in a persistent store to recall them on any scenario or campaign at the moment. _If_ you ever run into plain id conflicts you can always just generate your own UUIDs in e.g. Lua.
20131003 00:53:22< justinzane> The looking I've done, it seems like everything ends up getting persisted as a std::string since everything goes out to and comes in from WML. So, having a serial uint32 id does not seem to matter. I shall go off a digging, I shall...
20131003 00:53:27< shadowm> Of course this kind of thing is generally discouraged in normal gameplay because it absolutely ruins balance.
20131003 00:54:22< shadowm> WML attributes since 1.9.x actually have an underlying variant type.
20131003 00:55:32< shadowm> So if the game code (or Lua) wants a WML attribute to be one of the supported types, it'll get retrieved from memory and saved to memory (and written to disk) with that type.
20131003 00:57:12< anonymissimus> gfgtdf: what's the boost version you're using in your build setup ?
20131003 00:57:13< gfgtdf> anonymissimus: shadowm	dais it could also be due to different boost versions i use boost 1.51
20131003 00:57:26< gfgtdf> ^
20131003 00:57:31< shadowm> This results in boolean attributes being always stored to disk with unquoted yes/no values, et cetera. Since it is written to disk by the WML writer, a quick look at the preferences file should reveal whether a given attribute is internally handled as a string type or not.
20131003 00:57:42< anonymissimus> I seem to recall that I perhaps started to have those issues after changing my boost version
20131003 00:57:50< shadowm> Id est, all string attributes are written to disk in quoted form.
20131003 00:57:59< anonymissimus> that is, the cache and save index crashes
20131003 00:58:41< gfgtdf> anonymissimus. what version do you use?
20131003 00:58:54< anonymissimus> 1_46_1
20131003 00:59:23< anonymissimus> you can try 1.42 IIRC that was my previous one without those crashes
20131003 01:00:06< gfgtdf> :( i dot't realy want to downgrade my boost version.
20131003 01:00:45< anonymissimus> but it may also be that I didn't have them because the cache was never used until spring 1012 or so when a bug got fixed
20131003 01:00:53< anonymissimus> 2012
20131003 01:01:57< gfgtdf> anonymissimus: have you seen my pull request?
20131003 01:02:55-!- anonymissimus_ [~chatzilla@HSI-KBW-149-172-228-192.hsi13.kabel-badenwuerttemberg.de] has joined #wesnoth-dev
20131003 01:03:52< anonymissimus_> gfgtdf: if it gets merged, make MSC_VER >= 1500 then
20131003 01:05:03< gfgtdf> anonymissimus_: can you merge it then?
20131003 01:05:20< shadowm> The same bug I linked.
20131003 01:05:20< shadowm> No, the other one I didn't link, #19681.
20131003 01:05:30-!- anonymissimus [~chatzilla@HSI-KBW-149-172-228-192.hsi13.kabel-badenwuerttemberg.de] has quit [Ping timeout: 264 seconds]
20131003 01:05:30-!- Smar [smar@freenet/translator/finnish/Smar] has quit [Ping timeout: 264 seconds]
20131003 01:05:35-!- anonymissimus_ is now known as anonymissimus
20131003 01:05:38-!- Smar [smar@freenet/translator/finnish/Smar] has joined #wesnoth-dev
20131003 01:06:31< shadowm> 20:05:20 <anonymissimus_> gfgtdf: if it gets merged, make MSC_VER >= 1500 then
20131003 01:06:44< shadowm> Wouldn't one do that _before_ merging? :/
20131003 01:13:05< _Coffee> seen thunderstruck?
20131003 01:13:21< shadowm> wesbot: seen thunderstruck
20131003 01:13:22< wesbot> shadowm: The person with the nick thunderstruck last spoke 9h 20m ago.  2h 15m ago was here and on the channel #wesnoth with the message: Quit: Leaving
20131003 01:13:22< anonymissimus> gfgtdf: I don't merge that
20131003 01:14:18< anonymissimus> (not because I think it's bad or so but because I cannot motivate me to do so)
20131003 01:16:41< shadowm> Should we sing a song to motivate you?
20131003 01:17:11< gfgtdf> anonymissimus: i thought it is only pressing the "merge" button?
20131003 01:17:47< shadowm> I am pretty sure the process is usually more involved than that. Reviewing patches, making sure they don't break anything, et cetera.
20131003 01:18:08< shadowm> At least that's how I have always handled patches.
20131003 01:18:20-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Remote host closed the connection]
20131003 01:19:03< gfgtdf> hm ok but unlike my last patch that should be quiet easy this time.
20131003 01:21:46< shadowm> Why are "msvc fix: catching gzip_error when reading cache" and "added include for gzip_error" separate commits?
20131003 01:23:32< shadowm> Also, I'd personally prefer if there weren't so many typos in the commit messages and they were generally more descriptive of what they are doing.
20131003 01:23:52< gfgtdf> shadowm: because i forgot to include the header in the first commit
20131003 01:24:07< shadowm> Taking a look at this, for example: https://github.com/gfgtdf/wesnoth-old/commit/73e33930b29421c60af96167b658356719014715
20131003 01:24:24< shadowm> So, okay, there's a naming conflict. What header? Under what conditions? What were the consequences? Et cetera.
20131003 01:25:33< shadowm> This: https://github.com/gfgtdf/wesnoth-old/commit/b98aadd92cbb309008d3f8b801ec89929498b416  Notice what happens when your commit subject line is too long.
20131003 01:26:15< gfgtdf> hm i dont see a problem this that commit message.
20131003 01:26:16< shadowm> (As an aside, since I'm not a C++ lawyer I'm curious: why static_cast and not reinterpret_cast? The cast target is a pointer type, after all.)
20131003 01:26:32< shadowm> gfgtdf: You don't see it broken into two lines?
20131003 01:26:52< shadowm> The first line ends at "detai" and the second begins with "ls/520043/".
20131003 01:27:22< gfgtdf> hm ye i do but i can still click the link corrctly. and i dont think people would want to read the complete link anyway.
20131003 01:27:31< shadowm> Also, expecting me to follow the link while inspecting the patch list feels incredibly lazy.
20131003 01:27:41< shadowm> You are supposed to expect me to be lazy as well. :p
20131003 01:27:51< shadowm> No, I actually want to know what's going on in the commit.
20131003 01:28:44< shadowm> That's how _I_ handle patches. Some of the other people here just click on merge and expect the best, but I make sure to know what the potential side-effects are.
20131003 01:29:57< shadowm> So yeah, I'm a pedant because I don't want to be responsible for deploying nasty surprises in production. That's why I usually do not handle patches and pull requests. :p
20131003 01:31:38< shadowm> Having readable commit messages is also extremely important in the event that I need to figure out later why something isn't working correctly or why something was done in a certain way.
20131003 01:33:32< gfgtdf> well  dont think i could have written it shorter/better than the first 4 lines in the link, so i just put the link. and i dont think that is more work for the reader.
20131003 01:35:04< shadowm> "MSVC++ fix: explicitly cast NULL to pointer type \n <link>"
20131003 01:35:26< shadowm> I mean \n\n
20131003 01:36:23< shadowm> Even better, "MSVC++ 2010 fix: explicitly cast NULL to pointer type \n <link>".
20131003 01:36:30< shadowm> So yeah, that's an example.
20131003 01:36:37< gfgtdf> shadowm: hm i thought that was abvious from the code. people who want to know why something was done this way anready have ready the code.
20131003 01:36:57< irker394> wesnoth: anonymissimus wesnoth-old:master 40d81d1e258c / projectfiles/CodeBlocks/wesnoth.cbp: cb project update http://git.io/mbdaVA
20131003 01:36:57< shadowm> It's not obvious if I'm looking at git log without the -p switch.
20131003 01:38:19< gfgtdf> for soneone who uses git log i thought "msvc bugfix" is enough information.
20131003 01:38:50< shadowm> Hence I said I'm being a pedant.
20131003 01:39:17< gfgtdf> i dont know that word.
20131003 01:40:02< shadowm> Well, that's why free dictionaries exist. ;)
20131003 01:40:24< anonymissimus> you should, as I wan every wesnoth dev to be one :)
20131003 01:41:26< shadowm> There's only a certain permissible amount of pedantry before it begins to hinder communication and productivity.
20131003 01:41:33< gfgtdf> well teh dictonaries say pedant = pedant in german so i still dont know
20131003 01:41:39< anonymissimus> the by far most severe work about merging a pull request/applying a patch is to take the responsibility for any bugs or problems (even in the far future) that might come out of that
20131003 01:42:15< shadowm> gfgtdf: In my experience when learning English, it is also useful to look up words in an English dictionary (not an English-to-<language> dictionary).
20131003 01:43:07< shadowm> Many words are far more nuanced than what a translation dictionary is willing to specify.
20131003 01:43:46< shadowm> Especially so with words from colloquial language.
20131003 01:44:55< shadowm> anonymissimus: I'd say the exact author:committer responsibility ratio varies depending on the nature of the patch(es) and the problems caused.
20131003 01:45:35< shadowm> If it's a simple syntax error in a one-liner, I'd assign 95% of the responsibility on the committer.
20131003 01:46:53< anonymissimus> I'm more concerned about the author going MIA
20131003 01:46:57< shadowm> On the other hand, if it's a memory access violation taking place under specific conditions on an unusual platform in code that's buried deep in a ~ 500 lines patch series, the committer's responsibility should be relatively low.
20131003 01:47:05< anonymissimus> (and the committer as well)
20131003 01:47:07< shadowm> 40%, tops.
20131003 01:48:10< shadowm> anonymissimus: Oh yeah, that's why it's always convenient to have at least two people on board who understand the code.
20131003 01:48:23< shadowm> Not... zero, like it's often the case around here.
20131003 01:48:40-!- vultraz [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20131003 01:48:40-!- vultraz [~chatzilla@124.109.10.167] has quit [Changing host]
20131003 01:48:40-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20131003 01:49:45< shadowm> So, about https://github.com/gfgtdf/wesnoth-old/commit/f663ec9718b9604f994c0e1264c23fc832fef1ca ...
20131003 01:50:18< gfgtdf> what is about that ?
20131003 01:50:34< shadowm> Does the comment in https://gna.org/bugs/?16544#comment14 apply to that code too?
20131003 01:51:15< gfgtdf> well i have updated the comment on  the lates commit of that pull request.
20131003 01:51:39< gfgtdf> that function already throws erros since parser(cfg, filter,validator)(); thors errors
20131003 01:51:46< gfgtdf> throws*
20131003 01:52:30< shadowm> +  // i still dont know how other cimpiler handle it. 
20131003 01:53:16< gfgtdf> yes that was teh question i asked you at teh beginnign of this conversation "teh game cache produces corruptes 2-bytes long .gz files, windows crashes when reading them, but other version not do you know why not "
20131003 01:53:27< shadowm> "I still don't know how other compilers handle it." :(
20131003 01:53:45< gfgtdf> what makes you so unhappy?
20131003 01:53:52< shadowm> Typos, grammar, punctuation!
20131003 01:54:16-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20131003 01:54:18< shadowm> And all that aside, it feels to me more like material for a completely separate pull request dealing with runtime issues as opposed to build-time issues.
20131003 01:54:46< shadowm> Especially since it contains an statement to the effect of "I don't know for sure what's going on here".
20131003 01:55:13< gfgtdf> shadowm: well i know what goin one there when compiles with visual studio.
20131003 01:55:35< gfgtdf> thats whay i put the #if there
20131003 01:55:53< shadowm> "this seems to fix an issue" "based on a guess" "i dont really know why"
20131003 01:56:19< gfgtdf> shadowm	: taht was a day ago now i know more than yestarday
20131003 01:56:31< shadowm> It says 35 minutes ago.
20131003 01:56:35< shadowm> https://github.com/gfgtdf/wesnoth-old/commit/063822d2718aa8128c6b078cfac72614942a33f3
20131003 01:57:13< gfgtdf> the later 2.
20131003 01:57:25< shadowm> It is one of the last two commits.
20131003 01:57:32< gfgtdf> oops forget to delete that line
20131003 01:57:40< shadowm> Yeah, why don't we do the following:
20131003 01:58:34< shadowm> 1) Create a new pull request dedicated solely to build-time and other trivial issues (including the symbol naming conflicts thing).
20131003 01:59:13< shadowm> 2) Create another pull request dedicated solely to changes that explicitly change behavior, like the aforementioned commit.
20131003 01:59:55< shadowm> I for one could merge (1), but not (2).
20131003 02:01:13< shadowm> And if you could make sure to run your comments and commit messages through a spellchecker that'd be great too. Another thing is that some of us generally frown upon comments that implicitly "own" code by speaking in the first person ("I", "me", "mine", etc.).
20131003 02:02:50< shadowm> Also, why is this labeled a "msvc fix"? https://github.com/gfgtdf/wesnoth-old/commit/5dcc0588b0c69c99739bd5a277b2123c922d0800 I would expect that exception to be a valid candidate for catching there regardless of the compiler in use.
20131003 02:02:52< gfgtdf> well that wenoth compiles with visual studio crashes nearly every time you start it is a fact and i dont have those crashes so even if that code has bugs (what i dont think) having it is always beter than not having it.
20131003 02:03:31< Turuk> wesbot: seen zookeeper
20131003 02:03:31< wesbot> Turuk: The person with the nick zookeeper last spoke 2h 32m ago.  2h 12m ago was here and on the channels #wesnoth and #wesnoth-umc-dev with the message: Ping timeout: 240 seconds
20131003 02:03:34< gfgtdf> well i wrote that code that doesnt mean i own it, but can cannot speak for other people
20131003 02:04:03< shadowm> I wasn't speaking from a technical standpoint, only about the language employed there.
20131003 02:04:25< shadowm> "Who wrote this? The name that is in the file's copyright line?"
20131003 02:04:49< shadowm> Thinking of the case that I'm reading the code from the release tarball and not the git repository.
20131003 02:05:20< gfgtdf> the other compiler dot seem to need it thats why i wrote " i still dont know how other compiler handle it. ".  thast why i call it a msvc fix
20131003 02:05:29< shadowm> Also, I didn't say "no, we can't have this code", I'm only recommending better commit organization to speed up the merging process.
20131003 02:05:49< gfgtdf> s/dot/don't
20131003 02:06:16< anonymissimus> just add the nick to the I or me so it's easier to blame someone and I'm happy ;)
20131003 02:06:18< shadowm> If you don't want to fulfill requirements I won't be able to merge the whole thing myself since the commits that fall under the second proposed request are well beyond my grasp.
20131003 02:06:23< shadowm> *my requirements
20131003 02:07:26< gfgtdf> you normaly can easily press "blame" on github to see who wrote that, but i can add that.
20131003 02:09:34< gfgtdf> i DID remove that line saying it is based on a guess.
20131003 02:09:45< gfgtdf> i dont know what it wasnt displaying to you
20131003 02:10:15< shadowm> Ah, the very last commit, right.
20131003 02:14:02-!- Octalot [~noct@host86-186-158-121.range86-186.btcentralplus.com] has quit []
20131003 02:15:22-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20131003 02:19:08-!- Gallaecio [~quassel@84.120.112.34.dyn.user.ono.com] has quit [Ping timeout: 240 seconds]
20131003 02:22:13< mattsc> Hey all.  I need some C++ help/advice.  I need to get this bug fixed: https://gna.org/bugs/?func=detailitem&item_id=19769
20131003 02:22:34< mattsc> It's caused by this line:  https://github.com/wesnoth/wesnoth-old/blob/master/src/ai/contexts.cpp#L1305
20131003 02:23:17< mattsc> not working, because time_of_day_variable.name() cannot be compared directly to std::string  (I'm not using the right lingo here, but I hope it makes sense)
20131003 02:23:56< shadowm> You mean resources::tod_manager->get_time_of_day().name ?
20131003 02:23:58< mattsc> I have a working solution, but I have no idea whether it includes something I shouldn't do (remember that I am not a C++ coder)
20131003 02:24:24< mattsc> shadowm: yes, the comparison with the elements of 'times' always returns 0
20131003 02:24:33-!- Turuk [~Turuk@cpe-65-29-217-233.cinci.res.rr.com] has quit [Quit: Turuk]
20131003 02:24:33< mattsc> even if the two strings are identical
20131003 02:25:03< mattsc> This is my solution that works, at least on my computer: http://pastebin.com/jaPEBRXi
20131003 02:25:49< shadowm> Are you sure it always returns false? That's odd.
20131003 02:25:56< shadowm> I mean 0.
20131003 02:26:07< shadowm> It's operator==() that returns false.
20131003 02:26:09< mattsc> Yes, I put a lot of debug output in there and that's what it did
20131003 02:26:50< shadowm> Your replacement code should work okay, although you can also make it a const reference like this:
20131003 02:26:54< shadowm> const std::string& tod_string = resources::tod_manager->get_time_of_day().name;
20131003 02:27:41< shadowm> You might want to mention in a comment and the commit why you need the tod_string variable, in case someone ever decides to try to optimize that and inadvertently undo the fix.
20131003 02:27:50< shadowm> *undoes
20131003 02:28:50< mattsc> shadowm: okay, thanks, I will try out your version later tonight, add the comment and then commit it.
20131003 02:29:49-!- Gallaecio [~quassel@84.120.112.34.dyn.user.ono.com] has joined #wesnoth-dev
20131003 02:31:03< shadowm> I'm not really sure why the original fails in the first place, though.
20131003 02:31:31< shadowm> operator const std::string&() const { return get(); }
20131003 02:31:34< shadowm> bool operator==(const std::string& o) const { return get() == o; }
20131003 02:31:43-!- anonymissimus [~chatzilla@HSI-KBW-149-172-228-192.hsi13.kabel-badenwuerttemberg.de] has quit [Quit: ChatZilla 0.9.90.1 [Firefox 11.0/20120312181643]]
20131003 02:32:12< shadowm> So theoretically both the current version of the code and yours are functionally identical.
20131003 02:32:33< mattsc> shadowm: neither am I, but it definitely doesn't work and it's not just restricted to OS X (the bug report says Fedora 16)
20131003 02:32:49< mattsc> If you want to test it yourself, I can give you a very simple test case.
20131003 02:33:06< shadowm> Sure.
20131003 02:35:07< mattsc> shadowm: put this http://pastebin.com/7XQQ88D1 into the Side 2 definition of the first scenario SotBE  (or of an AI side of any scenario really)
20131003 02:35:15-!- LordBob_ [~LordBob_@2a01:e34:ee82:47e0:21e:c2ff:fe01:261f] has left #wesnoth-dev []
20131003 02:35:50< mattsc> Start the campaign, and do :inspect -> Side 2 -> AI overview and look at the value of aggression
20131003 02:36:23< mattsc> It should change to 0.123 during morning, and go back to default at all other TODs
20131003 02:36:38< mattsc> It does with my modified version, it does not with the original, at least not for me.
20131003 02:39:35< shadowm> "Morning"?
20131003 02:39:44< shadowm> Isn't that supposed to be lowercase 'morning'?
20131003 02:40:09< mattsc> shadowm: no, if you look at that line of code, it actually compares the name, not the id
20131003 02:40:35< shadowm> That's not good. :\
20131003 02:40:40< mattsc> I printed out resources::tod_manager->get_time_of_day().name and it definitely has a capital letter at the beginning
20131003 02:40:45< shadowm> The name may be translated to any language.
20131003 02:40:57< shadowm> Well, in this case it isn't.
20131003 02:40:58< mattsc> Yeah, good point ...
20131003 02:41:21< shadowm> I mean, yes, in this case it _might_ be, if the user has selected a different language.
20131003 02:41:50< mattsc> shadowm: right.
20131003 02:42:14< mattsc> shadowm: I need to be offline for ~30 min, will check back in after that.
20131003 02:42:25< shadowm> Okay.
20131003 02:42:38-!- mattsc [~mattsc@fw.hia.nrc.ca] has quit [Quit: Ciao]
20131003 02:46:17-!- nurupo|away is now known as nurupo
20131003 02:58:29-!- gfgtdf [~chatzilla@f054150203.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.90.1 [Firefox 22.0/20130618035212]]
20131003 03:01:15-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 248 seconds]
20131003 03:06:42-!- gfgtdf [~chatzilla@f054150203.adsl.alicedsl.de] has joined #wesnoth-dev
20131003 03:06:54-!- mattsc [~mattsc@154.20.32.246] has joined #wesnoth-dev
20131003 03:07:34< gfgtdf> shadowm: do you knwo weather "filter.peek() != EOF" does imply in gcc  or clang "filter.good()" if filter is an istream ?
20131003 03:07:37< gfgtdf> know*
20131003 03:07:39< mattsc> shadowm: I'm back  (but will be afk on and off for a while).  
20131003 03:09:55< shadowm> gfgtdf: I don't really have any useful experience working with iostreams.
20131003 03:12:46< gfgtdf> hm ok, becaus if thats was the case i could delete the "i dont know how other compilers hande it" line too.
20131003 03:13:12< shadowm> That's why I can't handle that part of the pull request and why I suggested splitting it.
20131003 03:13:44< shadowm> mattsc: It's 0.4 during both morning and dawn for me.
20131003 03:14:54< shadowm> Without your patch, that is.
20131003 03:15:48< shadowm> And works as intended with your patch.
20131003 03:16:05< shadowm> I don't get it. I'm using English. :/
20131003 03:20:02< shadowm> mattsc: Anyway, the fact that it's using the ToD name and not the id remains a design bug to be solved.
20131003 03:20:56< shadowm> You probably know better than me who to pester about it, though. :p
20131003 03:21:54< shadowm> Uhm, what are those X's I see on the gamemap?
20131003 03:23:42< shadowm> mattsc: A quick grep reveals that all other WML code is passing ToD ids, not names. Was this implementation change recent?
20131003 03:25:18< mattsc> shadowm: thanks for testing - I don't think this code has been touched recently, at least not that I am aware of.
20131003 03:25:45< shadowm> Okay, so I changed the WML to use the id instead of the name, and replaced the C++ to compare the current ToD id, not name, and that works as intended too.
20131003 03:25:49< mattsc> I agree that it should be switched to id's.  Since it wasn't working before anyway, I don't think it's a big deal to do this.
20131003 03:26:10< shadowm> It wasn't working?
20131003 03:26:31< mattsc> No, because of that bug?  (The setting of the aspect by ToD, I mean)
20131003 03:26:41< mattsc> s/?/.
20131003 03:27:20< mattsc> So wherever it is implemented in a campaign, it wouldn't have taken effect anyway.
20131003 03:27:28< mattsc> shadowm: do you just want to commit it?
20131003 03:27:43< mattsc> I'll go through mainline campaigns afterward and change the WML, if needed.
20131003 03:28:03< shadowm> It seems to have been introduced by Crab_ in f788542095254e636c65ebae7055f2126df69ef5 in 2009.
20131003 03:28:21< shadowm> mattsc: No, that's the matter. All WML in mainline passes ids to that attribute, not names.
20131003 03:28:37< mattsc> shadowm: oh, I misunderstood that.
20131003 03:28:46< shadowm> Which is the sane thing to do. I assume someone just mistook names for ids when writing the C++.
20131003 03:28:59< mattsc> shadowm: well, then ithat's a second reason why that wouldn't have worked.
20131003 03:29:30< shadowm> So, is this all I need to change?
20131003 03:29:54< mattsc> shadowm: yes, I believe so.  I'll go over things later and verify that.
20131003 03:30:11< mattsc> ... and change anything that needs to be changed as a result, I mean
20131003 03:30:30< mattsc> shadowm: and the x's are an artifact of flix' recruiting code.  They only show up in debug mode.  And flix has already said that he'll remove them before the next release.
20131003 03:30:32-!- fabi [~fabi@wesnoth/developer/fendrin] has joined #wesnoth-dev
20131003 03:30:51< mattsc> artifact = debugging info
20131003 03:31:17< shadowm> mattsc: Ah, okay, I was curious if it was a scenario thing.
20131003 03:31:24< justinzane> Are there plans to **require** (as opposed to support) c++11?
20131003 03:31:44< shadowm> So... this means that ToD-specific AI parameters haven't worked since 2009? :/
20131003 03:31:59< shadowm> I guess I'll have to replay my campaign after this commit.
20131003 03:33:10-!- Gallaecio_ [~quassel@84.120.112.34.dyn.user.ono.com] has joined #wesnoth-dev
20131003 03:34:04-!- fendrin_ [~fabi@88-134-45-105-dynip.superkabel.de] has quit [Ping timeout: 264 seconds]
20131003 03:34:11< shadowm> mattsc: time_of_day is what's called an "aspect" in AI configuration parlance, right?
20131003 03:34:23-!- Gallaecio [~quassel@84.120.112.34.dyn.user.ono.com] has quit [Ping timeout: 260 seconds]
20131003 03:34:47< mattsc> shadowm: yes, all of that is correct
20131003 03:35:10< mattsc> shadowm: this one here: http://wiki.wesnoth.org/AiWML#The_.5Bai.5D_Tag:_Defining_Aspects
20131003 03:36:04< mattsc> shadowm: and yes, my guess is that they have not worked in a long time.  A lot of the effects of AI aspects are pretty subtle, so they are hard to notice unless you really dig in.
20131003 03:37:52< shadowm> Yeah. :(
20131003 03:39:02< gfgtdf> justinzane: i asked that question before and if i remember correctly the answer was the we dont want to be "early adapters"
20131003 03:41:44< irker394> wesnoth: Ignacio R. Morelle wesnoth-old:master b7ad119de231 / changelog src/ai/contexts.cpp: Match time_of_day AI aspect against the current ToD id, not its name http://git.io/CorjJg
20131003 03:41:51< shadowm> mattsc: There.
20131003 03:42:51< mattsc> shadowm: thanks!  I'll check out a few things later tonight and then mark that bug as fixed too.
20131003 03:42:52< shadowm> Oh man I fixed an AI bug.
20131003 03:43:02< mattsc> shadowm: yay!  :)
20131003 03:43:09< shadowm> mattsc: OK, thank you for noticing though.
20131003 03:43:29< shadowm> Not many people seem to bother with tracking the AI configuration that closely.
20131003 03:43:53< mattsc> shadowm: no worries.  I actually have a use case for which I really need it to work, so it's a very selfish interest.
20131003 03:44:23< shadowm> That's how bugs get discovered though.
20131003 03:45:02< shadowm> There are approximately 83 instances of the aspect in mainline campaigns, so it's no minor thing.
20131003 03:45:07< mattsc> shadowm: a side effect might be that some scenarios might be playing out slightly differently now, if this suddenly starts working.  But I doubt that's it a huge effect.
20131003 03:45:21< mattsc> shadowm: yeah ...  (just thinking the same thing)
20131003 03:47:30-!- exciton_ [chuck-the-@89.208.169.104] has joined #wesnoth-dev
20131003 03:48:45-!- _8680_ [~8680@2002:4404:712c:0:ecb8:8cac:848c:647] has quit [Ping timeout: 245 seconds]
20131003 03:49:29-!- _8680_ [~8680@2002:4404:712c:0:8f5:b1ce:82d5:1c3d] has joined #wesnoth-dev
20131003 03:49:51-!- _8680_ is now known as Guest2963
20131003 03:50:00-!- exciton_ [chuck-the-@89.208.169.104] has quit [Read error: Connection reset by peer]
20131003 03:50:21-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has joined #wesnoth-dev
20131003 03:50:43-!- exciton [chuck-the-@89.208.169.104] has quit [Ping timeout: 260 seconds]
20131003 03:51:33-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has quit [Client Quit]
20131003 03:51:54< irker394> wesnoth: Ignacio R. Morelle wesnoth-old:master b1c8c1a04eb8 / RELEASE_NOTES: Add note about time_of_day AI aspect fix to RELEASE_NOTES http://git.io/zkI2sQ
20131003 03:53:02-!- exciton [chuck-the-@89.208.169.104] has joined #wesnoth-dev
20131003 03:56:04< irker394> wesnoth: Ignacio R. Morelle wesnoth-old:master e914175bf583 / RELEASE_NOTES: Add note about the Game Paths dialog to RELEASE_NOTES http://git.io/2QtY5w
20131003 03:57:43-!- exciton [chuck-the-@89.208.169.104] has quit [Ping timeout: 260 seconds]
20131003 03:58:01-!- exciton [chuck-the-@89.208.169.104] has joined #wesnoth-dev
20131003 04:08:09-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has joined #wesnoth-dev
20131003 04:08:57< justinzane> gfgtdf: Early adopter I get, but according to http://gcc.gnu.org/projects/cxx0x.html there is only one feature missing. And, gcc is already beginning to support c++1y http://gcc.gnu.org/projects/cxx1y.html
20131003 04:09:20< gfgtdf> justinzane: gcc is not the ony compiler
20131003 04:09:23< gfgtdf> only*
20131003 04:09:27< justinzane> All the same, whatever the project leads like, ...
20131003 04:10:19< justinzane> clang already supports all of c++11 http://clang.llvm.org/cxx_status.html
20131003 04:11:42< shadowm> For compiling on wesnoth.org we need to support gcc 4.4 at the very minimum right now, soon only 4.7.
20131003 04:12:04< justinzane> icc supports all but one feature on *nix's http://software.intel.com/en-us/articles/c0x-features-supported-by-intel-c-compiler
20131003 04:12:04< shadowm> I think we also have a server using gcc 3.4.
20131003 04:12:22< shadowm> Yeah, 3.4.6.
20131003 04:12:28< justinzane> Wow
20131003 04:12:46< justinzane> Is there a reason that noone has updated that?
20131003 04:14:16< shadowm> Shruuuuuuuuuuuug. :(
20131003 04:14:18< gfgtdf> im off bye
20131003 04:14:24-!- gfgtdf [~chatzilla@f054150203.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.90.1 [Firefox 22.0/20130618035212]]
20131003 04:15:46 * justinzane wondering whether cornbread will rise with yeast?  off to experiment...
20131003 04:21:20-!- Ivanovic_ [~ivanovic@x2f3e7ef.dyn.telefonica.de] has joined #wesnoth-dev
20131003 04:21:20-!- Ivanovic_ [~ivanovic@x2f3e7ef.dyn.telefonica.de] has quit [Changing host]
20131003 04:21:20-!- Ivanovic_ [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20131003 04:25:04-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Ping timeout: 264 seconds]
20131003 04:25:20-!- Ivanovic_ is now known as Ivanovic
20131003 04:25:35-!- nurupo is now known as nurupo|away
20131003 04:26:01-!- fabi [~fabi@wesnoth/developer/fendrin] has quit [Quit: Konversation terminated!]
20131003 04:27:41-!- fabi [~fabi@wesnoth/developer/fendrin] has joined #wesnoth-dev
20131003 04:55:32-!- Guest2963 [~8680@2002:4404:712c:0:8f5:b1ce:82d5:1c3d] has quit [Ping timeout: 260 seconds]
20131003 04:56:27-!- _8680_ [~8680@2002:4404:712c:0:68b8:2af2:4395:9fe7] has joined #wesnoth-dev
20131003 04:56:49-!- _8680_ is now known as Guest28920
20131003 05:11:38-!- Guest28920 [~8680@2002:4404:712c:0:68b8:2af2:4395:9fe7] has quit [Ping timeout: 264 seconds]
20131003 05:12:29-!- _8680__ [~8680@2002:4404:712c:0:d949:1eba:6221:35c8] has joined #wesnoth-dev
20131003 05:15:19< mattsc> shadowm: thanks you.  I have confirmed that your bug fix works for my test case also.
20131003 05:16:05< mattsc> I also looked through the usage in mainline campaigns, and the vast majority is making the AI more aggressive at night, and less so during the day (or the other way around), ..
20131003 05:16:40< mattsc> ... but there is also a case in EI, where the leader goal is supposed to be determined by ToD.  I will check out what that is all about.
20131003 05:17:41< mattsc> I'm going to clarify on the AiWML wiki page that it is the ToD id that is needed and mark the bug as fixed.
20131003 05:21:01-!- _8680__ [~8680@2002:4404:712c:0:d949:1eba:6221:35c8] has quit [Ping timeout: 245 seconds]
20131003 05:22:16-!- _8680__ [~8680@2002:4404:712c:0:969:283d:cd4e:d2ac] has joined #wesnoth-dev
20131003 05:22:35-!- Ggambit [~derek@wesnoth/developer/grickit] has quit [Remote host closed the connection]
20131003 05:26:36< shadowm> mattsc: You might want to mention there that the name was incorrectly used for versions 1.7.4 through 1.11.6.
20131003 05:27:54< mattsc> shadowm: FYI: https://gna.org/bugs/?func=detailitem&item_id=19769
20131003 05:28:40< mattsc> shadowm: I mentioned on the wiki that it was broken "for a long time" - I'll add the version numbers since you dug them out.  :)
20131003 05:30:43< mattsc> shadowm: done
20131003 05:32:41-!- _8680__ [~8680@2002:4404:712c:0:969:283d:cd4e:d2ac] has quit [Ping timeout: 245 seconds]
20131003 05:33:51-!- _8680_ [~8680@2002:4404:712c:0:2593:6d0a:efd8:98d9] has joined #wesnoth-dev
20131003 05:34:13-!- _8680_ is now known as Guest4893
20131003 05:48:00-!- EliDupree [~quassel@66-189-34-122.dhcp.oxfr.ma.charter.com] has quit [Ping timeout: 252 seconds]
20131003 06:07:54< mattsc> shadowm: https://github.com/wesnoth/wesnoth-old/blob/master/data/campaigns/Eastern_Invasion/scenarios/13_The_Drowned_Plains.cfg#L97
20131003 06:08:07< mattsc> I am pointing to l.97, which explains what the dragon is supposed to do.
20131003 06:08:40< mattsc> That did not work in 1.11.6 either (I just played through the campaign again recently), but it does now.  So you fixed that also.  :)
20131003 06:08:53< shadowm> Hm, I seem to recall someone mentioning the dragon bug before.
20131003 06:09:49< shadowm> https://gna.org/bugs/?func=detailitem&item_id=18356#comment2
20131003 06:10:25< shadowm> That's the most I could find using the built-in search in gna.org and the forums.
20131003 06:10:30< mattsc> shadowm: Woot!  You fixed two bugs with a grand total of 4 letters changed.  :)
20131003 06:10:46< mattsc> And possibly some others as well ...
20131003 06:10:54< mattsc> I'll add a comment to that bug as well.
20131003 06:13:43< mattsc> shadowm: done also.  I will check tomorrow if there are some other things that might have also been affected and/or fixed.  I am too lazy to do that today any more.
20131003 06:17:31< shadowm> Ugh, why can't it be trivial to get a set of SVN commits with their messages and diffs in separate files?
20131003 06:18:56< shadowm> Well, it's trivial for a range, but not for exporting individual commits of the range to patch files.
20131003 06:20:54< shadowm> Opted for specifying the full revision list by hand. `for r in 55982 56595 56596 56597 56598; do svn log --diff -r $r > patches/svn-r$r ; done`
20131003 06:24:46< shadowm> Hm, no way I was going to be able to apply all the patches without running into conflicts.
20131003 06:34:45-!- ancestral [~ancestral@174-20-209-41.mpls.qwest.net] has joined #wesnoth-dev
20131003 06:38:35< irker394> wesnoth: Nils Kneuper wesnoth-old:master b5503264bf2d / website/gettext.wesnoth.org/ (6 files in 2 dirs): moved repo over to git links at github http://git.io/LcJITA
20131003 06:38:37< irker394> wesnoth: Nils Kneuper wesnoth-old:master 025603c8fd33 / website/gettext.wesnoth.org/index.lang.php: attempt at fixing some issues with the update script http://git.io/v4oJMQ
20131003 06:38:39< irker394> wesnoth: Espreon wesnoth-old:master 94b726a6a77a / website/gettext.wesnoth.org/config.php: Added wesnoth-ai to the list of textdomains. http://git.io/xE7oZQ
20131003 06:38:41< irker394> wesnoth: Ignacio R. Morelle wesnoth-old:master df1d220042ff / website/gettext.wesnoth.org/ (functions.php index.extra.php index.lang.php index.php): gettext.wesnoth.org: Clean-up GET parameters for possible XSS exploits http://git.io/F7fnJA
20131003 06:39:53< shadowm> I love how Git lets me impersonate people.
20131003 06:52:08< shadowm> ./data/tools/wesnoth/wescamp.py -scampaigns.wesnoth.org  <--  pfffft, what a relic.
20131003 06:53:39-!- ancestral [~ancestral@174-20-209-41.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20131003 06:54:59-!- Ganrao [Ganrao@cpe-65-189-245-210.woh.res.rr.com] has quit [Ping timeout: 260 seconds]
20131003 06:59:23-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has quit [Quit: Leaving]
20131003 07:28:24-!- tomreyn_ [~tomreyn@megaglest/team/tomreyn] has joined #wesnoth-dev
20131003 07:29:07-!- exciton [chuck-the-@89.208.169.104] has quit [Ping timeout: 260 seconds]
20131003 07:29:34-!- bumbadadabum [~bumbadada@d155109.upc-d.chello.nl] has joined #wesnoth-dev
20131003 07:30:46-!- exciton [chuck-the-@89.208.169.104] has joined #wesnoth-dev
20131003 07:31:41-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has quit [Ping timeout: 245 seconds]
20131003 07:33:28-!- _Coffee [~david@ppp118-210-47-154.lns20.adl2.internode.on.net] has quit [Ping timeout: 240 seconds]
20131003 07:34:18-!- tomreyn_ is now known as tomreyn
20131003 07:38:31-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has quit [Quit: tomreyn]
20131003 07:38:35-!- ancestral [~ancestral@174-20-209-41.mpls.qwest.net] has joined #wesnoth-dev
20131003 07:45:40-!- mattsc [~mattsc@154.20.32.246] has quit [Quit: Ciao]
20131003 08:05:06-!- bumbadadabum [~bumbadada@d155109.upc-d.chello.nl] has quit [Ping timeout: 240 seconds]
20131003 08:21:53-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20131003 08:22:26< Rhonda> shadowm: Probably because they changed the HTML and the regex doesn't match anymore.  Thanks for the notice, will look into it.
20131003 08:23:08-!- trademark_ [~trademark@nsg93-8-88-175-59-164.fbx.proxad.net] has joined #wesnoth-dev
20131003 08:39:22-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20131003 09:05:01-!- boucman_work [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20131003 09:54:33-!- LordBob_ [~LordBob_@2a01:e34:ee82:47e0:21e:c2ff:fe01:261f] has joined #wesnoth-dev
20131003 10:11:11-!- bumbadadabum [~bumbadada@d155109.upc-d.chello.nl] has joined #wesnoth-dev
20131003 10:12:36-!- ancestral [~ancestral@174-20-209-41.mpls.qwest.net] has quit [Quit: And that’s the end of THAT chapter.]
20131003 10:16:52< irker394> wesnoth: Nils Kneuper website:master 1e2030cf203c / gettext.wesnoth.org/index.lang.php: attempt at fixing some issues with the update script http://git.io/8S4CAQ
20131003 10:16:54< irker394> wesnoth: Espreon website:master 0fd81f08f1c2 / gettext.wesnoth.org/config.php: Added wesnoth-ai to the list of textdomains. http://git.io/n7xTIA
20131003 10:16:56< irker394> wesnoth: Ignacio R. Morelle website:master 397136fd9c78 / gettext.wesnoth.org/ (functions.php index.extra.php index.lang.php index.php): gettext.wesnoth.org: Clean-up GET parameters for possible XSS exploits http://git.io/jaHJBA
20131003 10:17:42-!- mjs-de [~mjs-de@g224182236.adsl.alicedsl.de] has joined #wesnoth-dev
20131003 10:17:54 * shadowm is trying to understand how irker-webhook works.
20131003 10:21:08< shadowm> AI0867: Your run-tmux script sucks, it causes windows to be killed on CTRL-C.
20131003 10:24:30-!- H-Hour [~H-Hour@cpc7-sgyl35-2-0-cust428.18-2.cable.virginmedia.com] has joined #wesnoth-dev
20131003 10:25:11< shadowm> I mean tmux-start.
20131003 10:26:03< irker394> wesnoth: Ignacio R. Morelle website:master 4387bfd32674 / gettext.wesnoth.org/ (bin/update-gettext-stats config.php): gettext.w.o: Support for baldras.w.o instead of asheviere.w.o http://git.io/4ug_1g
20131003 10:26:06< irker394> wesnoth: Ignacio R. Morelle website:master b11e28e92a04 / gettext.wesnoth.org/bin/update-gettext-stats: gettext.wesnoth.org: Restore website repository auto-pull http://git.io/5qWGHA
20131003 10:26:09< irker394> wesnoth: Ignacio R. Morelle website:master 8c65297c332d / gettext.wesnoth.org/bin/update-gettext-stats: gettext.w.o: cd into gettext.w.o before running grab-stats http://git.io/za99HQ
20131003 10:26:12< irker394> wesnoth: Ignacio R. Morelle website:master ad9cdb840d3f / gettext.wesnoth.org/bin/update-gettext-stats: gettext.w.o: Don't require bash to run update-gettext-stats http://git.io/kT9W6g
20131003 10:26:59< shadowm> AI0867: Exaclty how do I change the "wesnoth:" prefix at the beginning? Or is that actually supposed to be the organization name?
20131003 10:28:12-!- Octalot [~noct@host86-159-75-217.range86-159.btcentralplus.com] has joined #wesnoth-dev
20131003 10:31:55< irker394> wesnoth: Ignacio R. Morelle website:master 56ebefd2b299 / gettext.wesnoth.org/index.php: gettext.w.o: Remove stray HTML comment in a PHP block http://git.io/VHzgjQ
20131003 10:33:06< irker394> wesnoth: Ignacio R. Morelle wesnoth-old:1.10 031463aeaa8a / / (525 files in 62 dirs): website: Delete tree, it's now split to a different repository http://git.io/9Ko2EA
20131003 10:34:27< irker394> wesnoth: Ignacio R. Morelle wesnoth-old:1.8 7d7c40151151 / / (433 files in 55 dirs): website: Delete tree, it's now split to a different repository http://git.io/1O-vZw
20131003 10:35:05< irker394> wesnoth: Ignacio R. Morelle wesnoth-old:1.6 19ffc61f1540 / website/start/ (169 files in 16 dirs): website: Delete tree, it's now split to a different repository http://git.io/crierw
20131003 10:35:39< irker394> wesnoth: Ignacio R. Morelle wesnoth-old:1.4 59d29f7628eb / website/start/1.4/ (8 files in 3 dirs): website: Delete tree, it's now split to a different repository http://git.io/Ji79BA
20131003 10:42:20-!- markus__ [~mjs-de@d184156.adsl.hansenet.de] has joined #wesnoth-dev
20131003 10:43:39-!- boucman_work1 [~rosen@193.56.60.161] has joined #wesnoth-dev
20131003 10:44:38-!- boucman_work [~rosen@wesnoth/developer/boucman] has quit [Ping timeout: 240 seconds]
20131003 10:46:03-!- mjs-de [~mjs-de@g224182236.adsl.alicedsl.de] has quit [Ping timeout: 260 seconds]
20131003 10:54:52< irker394> wesnoth: Ignacio R. Morelle website:master 5b2f86ab3d7c / gettext.wesnoth.org/config.php: gettext.w.o: Replace undefined $home variable with a valid string http://git.io/Ds407g
20131003 10:58:54< irker394> wesnoth: Ignacio R. Morelle website:master e39d559c30fd / gettext.wesnoth.org/bin/update-gettext-stats: gettext.w.o: Split CHECKOUTPATH into two differents vars http://git.io/XfWfwg
20131003 11:01:30< irker394> wesnoth: Ignacio R. Morelle website:master b2f86ea90f02 / gettext.wesnoth.org/bin/update-gettext-stats: gettext.w.o: Don't duplicate a path component from $WESCAMP_BASE http://git.io/lTdoMw
20131003 11:13:06-!- DCW [~Thunderbi@cpc1-finc14-2-0-cust12.4-2.cable.virginmedia.com] has joined #wesnoth-dev
20131003 11:15:05< shadowm> Hm, it seems 'localhost' on my system is resolving to ::1 when trying to connect to an add-ons server using the game client.
20131003 11:15:22< shadowm> And since it doesn't support IPv6, it fails with Broken pipe.
20131003 11:15:54< shadowm> No wait, that's not it.
20131003 11:16:21< shadowm> OK, I don't know why, but localhost fails and 127.0.0.1 works.
20131003 11:16:56< shadowm> It's not resolving to ::1 because that's a different error.
20131003 11:26:19< shadowm> fabi: I'm getting 20131003 06:25:32 warning unit: Unknown attribute 'base_ids' discarded.
20131003 11:26:31< shadowm> fabi: During an [object] [effect] apply_to=variation sequence.
20131003 11:26:52< shadowm> Also, is [object] [effect] apply_to=variation going to use id instead of name too in the future?
20131003 11:28:03< shadowm> It is probably worth noting that the unit in question has an initial variation set before the [object] is applied.
20131003 11:45:30-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 252 seconds]
20131003 11:54:55-!- DCW [~Thunderbi@cpc1-finc14-2-0-cust12.4-2.cable.virginmedia.com] has quit [Remote host closed the connection]
20131003 12:02:49-!- Ivanovic_ [~ivanovic@x2f48000.dyn.telefonica.de] has joined #wesnoth-dev
20131003 12:02:49-!- Ivanovic_ [~ivanovic@x2f48000.dyn.telefonica.de] has quit [Changing host]
20131003 12:02:49-!- Ivanovic_ [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20131003 12:05:52-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Ping timeout: 264 seconds]
20131003 12:06:47-!- Ivanovic_ is now known as Ivanovic
20131003 12:14:16< fabi> shadowm: hello
20131003 12:15:11< fabi> shadowm: Yes apply_to= needs to use id instead of the translated name.
20131003 12:22:36-!- lipkab [~the_new_l@2001:738:5404:192:915a:555:26a2:7028] has joined #wesnoth-dev
20131003 12:24:42-!- H-Hour_ [~H-Hour@cpc7-sgyl35-2-0-cust428.18-2.cable.virginmedia.com] has joined #wesnoth-dev
20131003 12:28:16-!- H-Hour [~H-Hour@cpc7-sgyl35-2-0-cust428.18-2.cable.virginmedia.com] has quit [Ping timeout: 246 seconds]
20131003 12:28:21-!- bumbadadabum [~bumbadada@d155109.upc-d.chello.nl] has quit [Ping timeout: 245 seconds]
20131003 12:33:21-!- markus__ is now known as mjs-de
20131003 12:46:07-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20131003 13:25:26-!- lipkab [~the_new_l@2001:738:5404:192:915a:555:26a2:7028] has quit [Quit: Távozom]
20131003 13:34:56-!- horon [~horon@nttkyo176024.tkyo.nt.ngn2.ppp.infoweb.ne.jp] has joined #wesnoth-dev
20131003 13:37:58-!- Octalot [~noct@host86-159-75-217.range86-159.btcentralplus.com] has quit []
20131003 13:38:44-!- Octalot [~noct@host86-159-75-217.range86-159.btcentralplus.com] has joined #wesnoth-dev
20131003 13:39:14-!- shadowm [~ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 264 seconds]
20131003 13:39:45-!- shadowm [~ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20131003 13:53:47-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has joined #wesnoth-dev
20131003 14:01:34-!- irker394 [~irker@ai0867.net] has quit [Quit: transmission timeout]
20131003 15:07:20-!- boucman_work1 is now known as boucman_work
20131003 15:07:37-!- boucman_work [~rosen@193.56.60.161] has quit [Changing host]
20131003 15:07:37-!- boucman_work [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20131003 15:08:59-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has quit [Quit: tomreyn]
20131003 15:16:16-!- Kostic [~marko@85.202.113.53] has joined #wesnoth-dev
20131003 15:19:21-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has joined #wesnoth-dev
20131003 15:45:10-!- mattsc [~mattsc@S010610bf4849b5d4.gv.shawcable.net] has joined #wesnoth-dev
20131003 15:52:30-!- Turuk [Turuk@cpe-65-29-217-233.cinci.res.rr.com] has joined #wesnoth-dev
20131003 15:52:57-!- Turuk is now known as Guest34637
20131003 16:02:17-!- EliDupree [~quassel@66-189-34-122.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20131003 16:15:18< Rhonda> hah
20131003 16:15:31< Rhonda> ERROR: The certificate of `gna.org' is not trusted.
20131003 16:15:31< Rhonda> The certificate has expired
20131003 16:16:43< Rhonda> wesbot: reload
20131003 16:16:43< wesbot> Rhonda: Yes, m'Lady!
20131003 16:16:46< Rhonda> wesbot: topic
20131003 16:16:47-!- wesbot changed the topic of #wesnoth-dev to: current repo: https://github.com/wesnoth/wesnoth-old (Please read the dev-ml for the latest status!) | 198 bugs, 339 feature requests, 28 patches | Logs: http://irclogs.wesnoth.org | Don't paste on IRC! Use a pastebin: http://pastebin.com | http://imagebin.org
20131003 16:17:11< Rhonda> shadowm: Actually, blame gna :)
20131003 16:17:30< Rhonda> wesbot: bug 21158
20131003 16:17:31< wesbot> Bug #21158  Assigned to: None  Status: None  Priority: 5 - Normal
20131003 16:17:31< wesbot> Summary: "gamemap" class doesn't handle empty files
20131003 16:17:31< wesbot> Original submission: The gamemap class which is located in map.?pp should hand
20131003 16:17:34< wesbot> le empty files by throwing "incorrect_map_format_error".Currently, it still 
20131003 16:17:37< wesbot> URL: http://gna.org/bugs/?21158
20131003 16:17:57-!- flix [~flix@37-5-10-145-dynip.superkabel.de] has joined #wesnoth-dev
20131003 16:18:09 * Rhonda . o O ( sed -i -e s/https/http/g ~/.irssi/scripts/wesbot.pl )
20131003 16:19:14< Rhonda> Actually, expired in august already, and noone complained yet??  sheesh
20131003 16:21:27< zookeeper> Rhonda, hasn't that been the case for forever now? at least firefox has been telling be that for... i dunno, at least over a year?
20131003 16:21:39< zookeeper> unless you're talking of a different certificate.
20131003 16:23:06< Rhonda> esr, could you take a look into the gna! ssl certificate?  Seemingly expired in august, and you are part of the gna admin team.  :)
20131003 16:23:44< Rhonda> zookeeper: the one on https://gna.org/ says "valid until 10.08.13"
20131003 16:24:28< Rhonda> Seems to need to get updated every half a year.
20131003 16:35:11-!- horon [~horon@nttkyo176024.tkyo.nt.ngn2.ppp.infoweb.ne.jp] has quit [Quit: Leaving...]
20131003 16:36:08-!- Guest34637 is now known as Turuk
20131003 16:37:17-!- wesbot [~wesbot@asteria.debian.or.at] has quit [Remote host closed the connection]
20131003 16:39:08-!- mattsc [~mattsc@S010610bf4849b5d4.gv.shawcable.net] has quit [Ping timeout: 240 seconds]
20131003 16:39:24-!- bumbadadabum [~bumbadada@d155109.upc-d.chello.nl] has joined #wesnoth-dev
20131003 16:41:09-!- wesbot [~wesbot@wesnoth/bot/wesbot] has joined #wesnoth-dev
20131003 16:41:32-!- mattsc [~mattsc@71-19-182-26.dedicated.allstream.net] has joined #wesnoth-dev
20131003 16:49:27-!- nurupo|away is now known as nurupo
20131003 16:50:49-!- nurupo is now known as nurupo|away
20131003 16:54:42-!- irker569 [~irker@ai0867.net] has joined #wesnoth-dev
20131003 16:54:42< irker569> wesnoth: mattsc wesnoth-old:1.10 91bbf9516cbd / data/campaigns/The_Rise_Of_Wesnoth/utils/trow-macros.cfg: TRoW S6 S22: make spawned unit rise on same hex as killed unit http://git.io/mFVQ0g
20131003 17:09:01< irker569> wesnoth: mattsc wesnoth-old:1.10 0919bc23f492 / changelog: Update changelog http://git.io/nrqFww
20131003 17:17:07-!- timotei_ [~timotei@wesnoth/developer/timotei] has quit [Ping timeout: 260 seconds]
20131003 17:18:00-!- timotei_ [~timotei@wesnoth/developer/timotei] has joined #wesnoth-dev
20131003 17:24:28-!- travis-ci [~travis-ci@ec2-184-73-76-130.compute-1.amazonaws.com] has joined #wesnoth-dev
20131003 17:24:28< travis-ci> [travis-ci] wesnoth/wesnoth-old#671 (1.10 - 91bbf95 : mattsc): The build passed.
20131003 17:24:28< travis-ci> [travis-ci] Build details : http://travis-ci.org/wesnoth/wesnoth-old/builds/12095832
20131003 17:24:28-!- travis-ci [~travis-ci@ec2-184-73-76-130.compute-1.amazonaws.com] has left #wesnoth-dev []
20131003 17:37:03-!- flix [~flix@37-5-10-145-dynip.superkabel.de] has quit [Ping timeout: 240 seconds]
20131003 17:38:18-!- travis-ci [~travis-ci@ec2-184-73-76-130.compute-1.amazonaws.com] has joined #wesnoth-dev
20131003 17:38:18< travis-ci> [travis-ci] wesnoth/wesnoth-old#672 (1.10 - 0919bc2 : mattsc): The build passed.
20131003 17:38:18< travis-ci> [travis-ci] Build details : http://travis-ci.org/wesnoth/wesnoth-old/builds/12096415
20131003 17:38:18-!- travis-ci [~travis-ci@ec2-184-73-76-130.compute-1.amazonaws.com] has left #wesnoth-dev []
20131003 17:40:35-!- DCW [~Thunderbi@cpc1-finc14-2-0-cust12.4-2.cable.virginmedia.com] has joined #wesnoth-dev
20131003 17:42:05-!- flix [~flix@37-5-10-145-dynip.superkabel.de] has joined #wesnoth-dev
20131003 17:50:01< flix> mattsc: hey!
20131003 17:50:19< mattsc> flix: hi  - I need to run off in about 5 min
20131003 17:51:18< flix> mattsc: okay, will you come back later?
20131003 17:51:32< mattsc> Yes.  In about 3 hours.
20131003 17:56:37-!- mattsc [~mattsc@71-19-182-26.dedicated.allstream.net] has quit [Quit: Computer's napping]
20131003 17:57:53-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20131003 18:03:04-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20131003 18:17:17-!- DCW [~Thunderbi@cpc1-finc14-2-0-cust12.4-2.cable.virginmedia.com] has quit [Remote host closed the connection]
20131003 18:18:43-!- boucman_work [~rosen@wesnoth/developer/boucman] has quit [Ping timeout: 260 seconds]
20131003 18:22:04-!- thunderstruck [~thunderst@cpc5-sgyl29-2-0-cust174.sgyl.cable.virginmedia.com] has joined #wesnoth-dev
20131003 18:26:39-!- Octalot [~noct@host86-159-75-217.range86-159.btcentralplus.com] has quit [Ping timeout: 260 seconds]
20131003 18:28:48-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20131003 18:54:42-!- trademark_ [~trademark@nsg93-8-88-175-59-164.fbx.proxad.net] has quit [Ping timeout: 264 seconds]
20131003 19:23:45-!- lipkab [~the_new_l@host-91-147-212-174.biatv.hu] has joined #wesnoth-dev
20131003 19:37:35-!- lipkab [~the_new_l@host-91-147-212-174.biatv.hu] has quit [Ping timeout: 260 seconds]
20131003 20:25:26-!- mattsc [~mattsc@fw.hia.nrc.ca] has joined #wesnoth-dev
20131003 20:33:35-!- Kostic [~marko@85.202.113.53] has quit [Ping timeout: 260 seconds]
20131003 20:42:36< mattsc> flix: hi, I'm back.
20131003 20:45:06< flix> mattsc: hey, do you have something to discuss? I have some small things... In an hour I'll meet a friend to test recruitment again.
20131003 20:46:26< mattsc> flix: if you agree, could you change the defaults for begin/end parameters to 1.5/1.1.  That's my current best guess for a good general setting, although a lot more testing is obviously required.
20131003 20:46:58< flix> Yes, I'll do it.
20131003 20:47:29-!- Guest4893 is now known as _8680_
20131003 20:48:12< mattsc> flix: unless you have a better suggestion for what they should be.  But I think that the current defaults are definitely too low.
20131003 20:48:40< mattsc> Besides that, I have a couple minor things that we could discuss now or later, whatever works better for you.
20131003 20:49:44< flix> mattsc: now is fine.
20131003 20:50:21< mattsc> flix: first, just a reminder that you were going to get rid of those crosses (except maybe when debug logging is enabled)
20131003 20:50:23-!- gfgtdf [~chatzilla@f054150203.adsl.alicedsl.de] has joined #wesnoth-dev
20131003 20:50:55< flix> mattsc: Yes, will do it.
20131003 20:51:18< mattsc> flix, then a couple questions concerning your choices of parameters:
20131003 20:51:29< mattsc> "The AI will ignore all those gold saving strategies when there is a LEADER_IN_DANGER. This happens when an enemy is near the leader (3 hexes). "
20131003 20:51:46< flix> oh, it's 5 now
20131003 20:52:00< mattsc> Why 3 hexes (or any fixed number)?  Wouldn't "within one move of the leader" (ignoring enemies) make more sense?
20131003 20:52:17< mattsc> one move for the unit in question, not one move of the leader.
20131003 20:54:39< mattsc> flix: the other question is why -1 for spend_all_gold defaults to start gold + 1?  Intuitively, I'd expect that to mean infinity.
20131003 20:54:49< mattsc> As in, it never kicks in.
20131003 20:55:17< flix> mattsc: you mean "if an enemy unit can reach the AI's leader within one move the AI's leader is in danger"?
20131003 20:55:26< mattsc> flix: yes
20131003 20:56:28< flix> mattsc: Yes, it was just easier to code the other way, (I will change it, should be not so hard).
20131003 20:59:40< flix> mattsc: About -1: As default I wanted that it kicks in *at some point*. And *at some point" has to be below the start gold. (So it doesn't kick in immediately). So I thought it is nice to set the default to start_gold + 1. Then I came up with this "-1".
20131003 21:01:29< mattsc> flix: One of the problems I see with that is if you have an AI in a scenario that you start with 0 gold, and then give some gold at a later time in order for it to start recruiting.
20131003 21:02:43< flix> mattsc: Oh, true... 
20131003 21:02:59< mattsc> flix: to be honest, I don't know why this condition needs to kick in by default at some point at all (having it as a parameter is a different question and that's good), but that's probably just because I have not thought about it much.
20131003 21:04:00< mattsc> flix: and my final question/request is for you to update the AiWML wiki page, please.
20131003 21:04:09< flix> mattsc: My intension was the question "why saving all the gold when not spending it eventually?".
20131003 21:04:28< mattsc> flix: I don't think you need to add to that page, but putting in the appropriate links would be good.
20131003 21:05:27< mattsc> flix: sure, but the counter question is: if I already have what I deem enough units on the field, why should I spend more gold (even if I have it)
20131003 21:05:47< flix> mattsc: do you think I should insert all the aspects there or write something like "For the aspects related to recruitment see here"
20131003 21:05:59< mattsc> flix: I am not saying that you change it, just that your default is unintuitive to me
20131003 21:06:11< mattsc> ... that you *should change it ...
20131003 21:06:39< mattsc> flix: yes, one or the other  ;)
20131003 21:06:50< mattsc> flix: in other word, I'd do whatever is less work
20131003 21:07:09-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20131003 21:09:05< flix> mattsc: I get your point, and actually I'm fine with defining "-1" to be "infinity". I just thought especially in MP-games it could be interesting when the AI starts a huge offensive wave... 
20131003 21:10:48< mattsc> flix: sure, I get that point, but I although think that the AI in an MP game will likely never have more than its starting gold.  Well, at least no against a human player.
20131003 21:12:09< mattsc> flix: but as I said, do with it as you want.  As long as it is well defined, it's okay.  The only real issue I can think of right now is the above example, which means that the AI behavior is different for AIs that start with full gold, and those that are assigned their gold during the scenario.
20131003 21:12:36< mattsc> flix: that was all from me.  For now.
20131003 21:13:21< mattsc> flix: unless if you want me to give you a couple examples why it is impossible to find an ideal value for the begin/end parameters that will always work.
20131003 21:17:58< Turuk> Impossible? Lies.
20131003 21:19:06< mattsc> Turuk: don't backpedal on me, I already had you convinced of this.  :D
20131003 21:19:14< flix> :D
20131003 21:19:38< flix> mattsc: You convinced me with the "-1". I'll change it
20131003 21:20:22< mattsc> flix: okay
20131003 21:20:29< flix> mattsc: I have some minor things too:
20131003 21:21:15< mattsc> flix: btw, I really like the configurability of all of this.  At the moment I am just trying to figure out what a good default for an "average setting" is, so that's what all these comments are about.
20131003 21:21:21< flix> mattsc: Is it appropriate to file a bug to find someone who can think about how to update wmllint for the new aspects?
20131003 21:21:52< flix> okay
20131003 21:22:28< mattsc> flix: umm, good question ...  I'm not sure if this should be a bug report or done otherwise.
20131003 21:22:42< flix> mattsc: because I never worked with wmllint, but I guess some changes are needed
20131003 21:23:09< mattsc> flix: yes, agreed  (and I have used it, but not done anything else with it)
20131003 21:24:06< flix> mattsc: You now someone to bug with this? Otherwise I'd just file a report.
20131003 21:24:21< flix> s/now/know
20131003 21:25:25< Turuk> mattsc: I would never backpedal, just dance around the answer. :P
20131003 21:26:09< mattsc> I vaguely remember shadowm discussing wmllint with somebody recently on this channel, but I forgot who it was.  You should be able to find it in the logs though (I would say it was within the last 2 weeks, or at most slightly more than that).
20131003 21:26:44< mattsc> Or check out the commit history of wmllint
20131003 21:26:49< mattsc> flix: ^
20131003 21:27:50< mattsc> Turuk: :P  Btw, haven't answered your last question yet: I haven't tested the more defensive night-time behavior yet since the bug got fixed.  I hope to get to that tonight.
20131003 21:28:27< flix> shadowm: Can you help with updating wmllint to work with the new aspects inside [ai], or do you know someone who could do it? I never worked with wmllint.
20131003 21:28:30< mattsc> Turuk: I'd say wait until flix has changed the defaults for gold saving, and then rebuild.
20131003 21:29:43< flix> Turuk: I'll commit it later today or tomorrow!
20131003 21:30:19< mattsc> Turuk: you'll have to add a couple new files to your local Xcode projectfile for that to work: src/strftime.[ch]pp, or update from the repo version and then reset the compiler and architecture values.
20131003 21:31:02< mattsc> (Those files have nothing to do with the changes you want, but they got added recently, so your build will fail unless you add them.)
20131003 21:35:51< Turuk> Okay, I'll remember that when I go to rebuild
20131003 21:36:17-!- zookeeper2 [~lmsnie@87-100-225-7.bb.dnainternet.fi] has joined #wesnoth-dev
20131003 21:37:24< mattsc> Turuk: all you have to do is right-click on the flie list in the left pane (which you get to by clicking on the leftern-most icon on the top, IIRC), and choose "Add files to Wesnoth".
20131003 21:37:31-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 240 seconds]
20131003 21:38:19< flix> mattsc: Annother thing: Do you know why Knalgan Alliance (and only them) have a recruitment_pattern set when starting a MP-Game?
20131003 21:38:54< mattsc> flix: umm, no.  Do they?
20131003 21:39:28< flix> mattsc: jep :) . I'll do a git blame...
20131003 21:41:19< mattsc> flix: yes, indeed.
20131003 21:43:05< flix> mattsc: it's there since 2007
20131003 21:43:53< mattsc> flix: interesting.  I didn't even know that.
20131003 21:44:13< mattsc> flix: btw, I also like what you have done to the :inspect output in 'ai overview'
20131003 21:45:16< flix> mattsc: :) thanks
20131003 21:45:35-!- {V} [~V@139-79-ftth.on.nl] has quit [Read error: Connection reset by peer]
20131003 21:46:05-!- {V} [~V@139-79-ftth.on.nl] has joined #wesnoth-dev
20131003 21:46:19< shadowm> flix: Elvish_Hunter and groggy have hacked on wmllint lately. I know some Python but I'm not very familiarized with the wmllint codebase beyond the simplest plain-text substitutions. I'm also notably busy with the servers as of late. :/
20131003 21:47:17< flix> mattsc: It seems that there was once a recruitment_pattern for all factions, but only knalgans-default.cfg got updated to a new syntax.
20131003 21:47:41-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20131003 21:47:55< mattsc> flix: interesting
20131003 21:48:14< flix> shadowm: Thanks! I'll ask them
20131003 21:48:23< mattsc> flix: with your new recruiting, you can probably get rid of that for Knalgans also, right?
20131003 21:49:12< mattsc> flix: as in, I'd assume your new recruiter to do better than a fixed recruitment pattern?
20131003 21:49:23< flix> mattsc: Yes, I guess I will do it. I haven't any reason in mind why we should keep it like this
20131003 21:49:31< flix> mattsc: yes
20131003 21:49:58< shadowm> fabi: So, could you reproduce the base_ids warning? A simple grep should reveal the site that's generating it.
20131003 21:50:16< shadowm> It's somewhere in unit.cpp.
20131003 21:54:16< shadowm> mattsc: Most of the placement options advertised in the wiki have never been implemented in the first place, that's why map_overwrite doesn't 'work' -- it simply isn't a thing.
20131003 21:54:59< shadowm> Although that may have changed later since I see this now.
20131003 21:55:00< shadowm> src/actions/create.cpp:162:                     if ( place != "map_overwrite" ) {
20131003 21:55:32< mattsc> shadowm: Oh, I see.  Well, it does seem to work in 1.11.6+dev, but not in 1.10.6.
20131003 21:56:13< mattsc> shadowm: either way, that WML code takes care of the problem.
20131003 21:56:25< mattsc> *change to the WML code
20131003 21:56:46< shadowm> It looks like it was stealthily added by a refactoring commit.
20131003 21:58:07< shadowm> A git log -p search on the full master history indicates that it first appeared in 724618c28798b78b9101abd560d2db9c5bc35b83 .
20131003 21:59:11< shadowm> jamit: Is this conclusion correct?
20131003 22:00:02< shadowm> For convenient reference, that's version 1.11.1 (more specifically a few hours after 1.11.0).
20131003 22:01:56< flix> mattsc: Another thing: I want to change the score of the CA "MOVE_LEADER_TO_KEEP" to 190000, so this CA gets executed before the recruitment CA. My CA will handle multiple leaders better then. I will ask Crab before I do it, but can you think of something why I should not change it?
20131003 22:02:04-!- H-Hour_ [~H-Hour@cpc7-sgyl35-2-0-cust428.18-2.cable.virginmedia.com] has quit [Ping timeout: 248 seconds]
20131003 22:02:36-!- Turuk_ [~Turuk@cpe-65-29-217-233.cinci.res.rr.com] has joined #wesnoth-dev
20131003 22:03:22< mattsc> flix: does your new recruitment CA actually move the leaders around?
20131003 22:03:40< flix> mattsc: No
20131003 22:04:36< mattsc> flix: then I need more information why this is needed.  If the leader is not on the keep, no recruiting will happen (recruiting score is 0).  Then, move_leader_to_keep kicks in.  Then recruiting happens.  What am I missing here?
20131003 22:06:12-!- Turuk_ [~Turuk@cpe-65-29-217-233.cinci.res.rr.com] has quit [Client Quit]
20131003 22:06:47< flix> mattsc: When the AI has two leaders, one of them is on a keep: evaluate won't return 0 then and my CA will recruit with this one leader only. It would be better to move the second leader to the keep as well before recruitment starts, so my AI will recruit with both leaders equally.
20131003 22:07:30< mattsc> flix: oh, assuming that there are 2 keep hexes?
20131003 22:07:46< flix> yes
20131003 22:07:59-!- LordBob_ [~LordBob_@2a01:e34:ee82:47e0:21e:c2ff:fe01:261f] has quit [Quit: LordBob_]
20131003 22:08:13< flix> mattsc: or even more then one castle
20131003 22:09:39< mattsc> flix: but this latter case only matters if the AI does not have the gold for filling both hexes with recruits (or does not want to), right?
20131003 22:10:38< mattsc> flix: I think if you move move_leader_to_keep up, it will mess up move_leader_to_goals then ...
20131003 22:11:08< mattsc> The point of the latter is to have a leader go somewhere without returning to the keep, but recruit first on the first move.
20131003 22:11:29< mattsc> (it's score is in between recruitment and move_leader_to_keep)
20131003 22:11:34< mattsc> s/it's/its
20131003 22:13:22-!- H-Hour_ [~H-Hour@cpc7-sgyl35-2-0-cust428.18-2.cable.virginmedia.com] has joined #wesnoth-dev
20131003 22:14:46< flix> mattsc: I see...
20131003 22:17:38< flix> mattsc: Is *deactivating MOVE_LEADER_TO_KEEP CA, if the leader has a goal* a option/solution?
20131003 22:19:03< mattsc> flix: hmm, I think that would work.  I cannot come up with a reason why move_leader_to_keep would be needed (or even make sense) if a leader goal is set.
20131003 22:19:27-!- H-Hour_ [~H-Hour@cpc7-sgyl35-2-0-cust428.18-2.cable.virginmedia.com] has quit [Quit: Leaving]
20131003 22:21:46< flix> mattsc: Okay, I'll ask Crab, thanks so far
20131003 22:22:11< mattsc> flix: sounds good.  Thanks for all the work you're doing on this!
20131003 22:22:29-!- mattsc [~mattsc@fw.hia.nrc.ca] has left #wesnoth-dev []
20131003 22:23:07-!- mattsc [~mattsc@fw.hia.nrc.ca] has joined #wesnoth-dev
20131003 22:28:22< shadowm> loonycyborg: Did you CC this person when replying here? https://mail.gna.org/public/wesnoth-dev/2013-10/msg00001.html
20131003 22:29:08< loonycyborg> hmm, only replied to ml
20131003 22:31:47< shadowm> He asked to be CC'd because he isn't subscribed.
20131003 22:37:02-!- Turuk [Turuk@cpe-65-29-217-233.cinci.res.rr.com] has quit [Ping timeout: 240 seconds]
20131003 22:41:01-!- wesbot changed the topic of #wesnoth-dev to: current repo: https://github.com/wesnoth/wesnoth-old (Please read the dev-ml for the latest status!) | 197 bugs, 339 feature requests, 28 patches | Logs: http://irclogs.wesnoth.org | Don't paste on IRC! Use a pastebin: http://pastebin.com | http://imagebin.org
20131003 23:01:36-!- Ggambit [~derek@wesnoth/developer/grickit] has joined #wesnoth-dev
20131003 23:04:28-!- thunderstruck [~thunderst@cpc5-sgyl29-2-0-cust174.sgyl.cable.virginmedia.com] has quit [Quit: Leaving]
20131003 23:16:37< AI0867> shadowm: that is indeed the org name
20131003 23:19:44< AI0867> 10:17  * shadowm is trying to understand how irker-webhook works. <-- I have a refactored and bitbucket-capable version around, but I haven't actually tested it yet
20131003 23:20:42< shadowm> These scripts in general sorely need a way to reload the configuration without being restarted.
20131003 23:20:46< shadowm> Say, SIGHUP.
20131003 23:21:53< AI0867> Rhonda: nobody noticed because most browsers don't trust the certificate's issuer, so the browsers complain anyway. I suppose nobody noticed that the subject of the complaint changed
20131003 23:23:22-!- zookeeper2 is now known as zookeeper
20131003 23:23:24-!- zookeeper [~lmsnie@87-100-225-7.bb.dnainternet.fi] has quit [Changing host]
20131003 23:23:24-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20131003 23:23:27-!- Turuk [~Turuk@208.102.36.12] has joined #wesnoth-dev
20131003 23:23:37< AI0867> shadowm: yes, that sounds like a nice feature to have
20131003 23:23:38< shadowm> I wouldn't have noticed because for that very reason I have an exception added for gna.org.
20131003 23:27:18-!- mjs-de [~mjs-de@d184156.adsl.hansenet.de] has quit [Remote host closed the connection]
20131003 23:30:03-!- gfgtdf_ [~chatzilla@f054146190.adsl.alicedsl.de] has joined #wesnoth-dev
20131003 23:31:34-!- gfgtdf [~chatzilla@f054150203.adsl.alicedsl.de] has quit [Ping timeout: 240 seconds]
20131003 23:31:42-!- gfgtdf_ is now known as gfgtdf
20131003 23:54:38-!- bumbadadabum [~bumbadada@d155109.upc-d.chello.nl] has quit [Quit: Bye]
20131003 23:55:35-!- gfgtdf [~chatzilla@f054146190.adsl.alicedsl.de] has quit [Ping timeout: 240 seconds]
--- Log closed Fri Oct 04 00:00:43 2013