--- Log opened Thu Aug 04 00:00:16 2016
20160804 00:02:45< irker340> wesnoth: loonycyborg wesnoth:master d5370aa38263 / src/server/game.cpp: Fix segfault caused by incorrect use of std::copy_if https://github.com/wesnoth/wesnoth/commit/d5370aa3826303127dee20c72f1c44aa27a37507
20160804 00:03:02< loonycyborg> shadowm: this should fix it
20160804 00:04:18< loonycyborg> hmm I wonder if original code didn't use back_insterter
20160804 00:04:25< loonycyborg> or I forgot it during porting :P
20160804 00:05:37< celmin> I'll have to check if that fixes the crash I've been getting in the server.
20160804 00:07:47< fabi_> hmmmm
20160804 00:09:07< fabi_> The line  {NAMED_GENERIC_UNIT 1 "Desert Archer" 24 27 "Eranor" ( _ "Eranor")} results in a female elf named "Eranor".
20160804 00:09:25-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160804 00:09:56< celmin> Are you sure?
20160804 00:10:00< celmin> Maybe it's random gender.
20160804 00:10:36< fabi_> The macro is working right.
20160804 00:10:45< fabi_> I just wonder that it was never discoverd.
20160804 00:11:32< fabi_> So my macro rewriting results in some bug fixing as a side effect.
20160804 00:12:11< celmin> What macro rewriting?
20160804 00:12:15< fabi_> celmin: Maybe all those unit macros should be redesigned.
20160804 00:12:45< celmin> Redesigning macros is a little questionable, though.
20160804 00:12:53< celmin> Depending on the details of course.
20160804 00:13:03< fabi_> I think NAMED_GENERIC_UNIT is more or less useless without a parameter for the gender.
20160804 00:13:28< fabi_> maybe if it would default to male 
20160804 00:13:55< fabi_> Anyway, I do not change the macros yet.
20160804 00:14:41< fabi_> I am just replacing some {Makrocall parameter1...} [+unit] ....
20160804 00:17:35-!- gfgtdf [~chatzilla@x4e368e00.dyn.telefonica.de] has joined #wesnoth-dev
20160804 00:17:43-!- Duthlet [~Duthlet@dslb-178-005-055-087.178.005.pools.vodafone-ip.de] has quit [Quit: leaving]
20160804 00:17:59< gfgtdf> fabi_: to make it easier to port them to your engine?
20160804 00:18:09< gfgtdf> fabi_: well there are soem genderless types.
20160804 00:18:40< celmin> He wants to avoid using [+tags] in mainline for some reason (and it so happens that it's not really necessary for those cases).
20160804 00:19:15< gfgtdf> loonycyborg: the original code didn't use std::copy_if so it didnt need use back_inserter eigher.
20160804 00:21:45-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160804 00:23:18< fabi_> gfgtdf: And removing all unbalanced wml from mainline. But celmin did nearly all of that already :-)
20160804 00:23:50< celmin> Did I?
20160804 00:24:00< fabi_> [foreach] implementation
20160804 00:24:08< celmin> Oh, i suppose.
20160804 00:24:51< gfgtdf> fabi_: but the wml is still using the unbalances {FOREACH} macro.
20160804 00:25:14< fabi_> gfgtdf: Not on my disc anymore. Just a few more minutes and some testing.
20160804 00:30:42< fabi_> celmin: So far all {FOREACH} I transformed did not modify the array while iterating so I only used [foreach] and no [for] so far.
20160804 00:31:03< celmin> Okay.
20160804 00:31:26< celmin> [for] is the drop-in substitute though (which is why I reimplemented {FOREACH} in terms of it)
20160804 00:31:27< fabi_> You sounded like you espected quite the opposite.
20160804 00:31:42< celmin> Nah, I had no particular expectations, just what I said above.
20160804 00:32:06< fabi_> would you prefer when I use [for] instead?
20160804 00:32:19< celmin> The "minimal" change would be making all of them [for], essentially just expanding the macro, but there's no particular reason to be minimal.
20160804 00:33:32< fabi_> The [foreach] consturct is slightly larger than the {FOREACH} one. But it is only 3 lines.
20160804 00:34:09< fabi_> And I like how [foreach] looks better.
20160804 00:35:46< fabi_> Okay Eranor is male now.
20160804 00:40:54< irker340> wesnoth: fendrin wesnoth:master 5a9f45ddba28 / data/campaigns/Under_the_Burning_Suns/scenarios/01_The_Morning_After.cfg: UtBS 01: Replace some calls to unit_utils macros with [+unit] by single macro ca https://github.com/wesnoth/wesnoth/commit/5a9f45ddba284f006d646bcdc0e9b41125b79b76
20160804 00:41:19< mattsc> Talking about macros, the behavior of the MESSAGE macro has been altered (as I mentioned earlier today).
20160804 00:41:51< mattsc> I have a vague recollection that we have talked about this before, but don’t remember when that was and what the conclusion was
20160804 00:42:01< mattsc> However, I do think that that needs to be fixed.
20160804 00:42:17 * mattsc goes into hiding again
20160804 00:42:27-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160804 00:54:18< vultraz> MESSAGE macro?
20160804 00:54:39< vultraz> That's deprecated 
20160804 00:58:11< mattsc> vultraz: since when?
20160804 00:58:17< vultraz> many releases
20160804 00:58:37< vultraz> i guess i just never got around to adding the appropriate notices for all of the stuff I deprecated
20160804 00:58:38< celmin> What was the MESSAGE macro?
20160804 00:58:47< vultraz> celmin: literally [message]
20160804 00:59:06< celmin> Like {MESSAGE speaker text}?
20160804 00:59:16< mattsc> nope
20160804 00:59:40< vultraz> MESSAGE SPEAKER_ID IMAGE CAPTION_TEXT MESSAGE_TEXT
20160804 00:59:59< celmin> …caption?
20160804 01:00:14< vultraz> yes
20160804 01:00:19< celmin> That really sounds like too many arguments.
20160804 01:00:19< vultraz> messages can have custom captions
20160804 01:00:38< celmin> mattsc: How'd the behaviour change BTW?
20160804 01:01:20< mattsc> 20160803 14:14:33< mattsc> hmm, MESSAGE macro with “” or () for the caption does not display the unit name any more 
20160804 01:01:44< celmin> Ah.
20160804 01:01:49< vultraz> sounds like a thing with the [message] implementation
20160804 01:01:57< celmin> That might be directly related to, yeah, the Lua [message] implementation.
20160804 01:02:02< vultraz> but either way, don't use MESSAGE, mattsc  
20160804 01:03:08< celmin> I actually think that what you describe is the proper behaviour when considered in terms of the WML tag that the macro expands to.
20160804 01:03:44< mattsc> vultraz: there is no mention in the changelog anywhere of MESSAGE being depracated
20160804 01:03:58< mattsc> or even deprecated
20160804 01:03:58< vultraz> These things don't get changelog entries
20160804 01:04:09< celmin> What are you talking about?
20160804 01:04:09< vultraz> I forgot to add the in-game notices
20160804 01:04:17< celmin> Obviously a deprecation should have a changelog entry.
20160804 01:04:18< mattsc> so where do you mention that one of the most-used macros has been deprecated?
20160804 01:04:35< celmin> If it has no changelog entry, let's just say it's deprecated in 1.13.6.
20160804 01:05:07< celmin> It may be highly-used, but it's not a very good macro in my opinion (due to the number of arguments).
20160804 01:05:19< mattsc> celmin: the point is not whether that is expected behavior for a tag, but that a macro that has been around for ages suddenly doesn’t work any more as before
20160804 01:05:41< mattsc> celmin: I am not arguing that point at all
20160804 01:05:46< celmin> I understand.
20160804 01:05:59< celmin> It's not good that the tag unexpectedly changed behaviour.
20160804 01:06:01< vultraz> It's been deprecated since Match 2015
20160804 01:06:29< mattsc> vultraz: can you show me where that is documented?
20160804 01:06:36< vultraz> It isn't :|
20160804 01:06:54< vultraz> I couldn't figure out where to put the notices and then totally forgot to add them
20160804 01:06:55< mattsc> then, how or where is it deprecated
20160804 01:06:55< vultraz> Blah
20160804 01:07:18< mattsc> vultraz: I am being intentional a pain in the ass here, btw  :)
20160804 01:07:23< vultraz> Yes you are :P
20160804 01:07:30< mattsc> uh, correct grammar as needed
20160804 01:07:43< vultraz> everything in here is deprecated https://github.com/wesnoth/wesnoth/blob/master/data/core/macros/deprecated-utils.cfg
20160804 01:07:45< mattsc> but seriously, this is (or at least was) a widely used macro
20160804 01:07:47< vultraz> or is supposed to be
20160804 01:07:59< mattsc> we cannot silently deprecate it or change its behavior
20160804 01:08:52< vultraz> I believe I was stumped because I didn't know where to put the deprecation notices for text macros
20160804 01:08:56< mattsc> deprecating it is one of of not wanting to deal with fixing the changed behavior, but it needs to be clear to people
20160804 01:09:09< mattsc> *one way of
20160804 01:09:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160804 01:10:22< irker340> wesnoth: fendrin wesnoth:master f321e716853c / data/campaigns/Sceptre_of_Fire/scenarios/8_The_Dragon.cfg: SoF 08: Removed some more [+unit]. https://github.com/wesnoth/wesnoth/commit/f321e716853c6bd632f4bf2a8672f907257521ed
20160804 01:10:45< vultraz> DEPRECATE seems like actionwml..
20160804 01:11:19< celmin> The deprecation macro is ActionWML, yes. (Also works in storyscreen.)
20160804 01:11:30< vultraz> but what about text macros?
20160804 01:11:35< celmin> What we really need is a #deprecated preprocessor instruction.
20160804 01:11:52< vultraz> Yes
20160804 01:11:58< mattsc> vultraz: I am using that macro 1,880 times in my add-ons  :P
20160804 01:12:14< vultraz> poor you :P
20160804 01:12:28< vultraz> you will be flooded with warnings
20160804 01:12:50< mattsc> Not because I think it’s great, or even good, in any way, but because that’s what I was doing back then when I started and never saw a reason to change it
20160804 01:12:55< vultraz> i suppose certain people will insist it persist until 1.15.
20160804 01:13:25< mattsc> and yes, of course I know how to deal with that without having to change it manually almost 2000 times 
20160804 01:13:40-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 258 seconds]
20160804 01:13:59< vultraz> if it were me I would delete them all right now
20160804 01:14:27< vultraz> but then again, I'm not the best role model :P
20160804 01:14:52-!- gfgtdf_ [~chatzilla@x4e369370.dyn.telefonica.de] has joined #wesnoth-dev
20160804 01:15:39< celmin> Since no-one knew it was deprecated, we need to somehow fix the issue of the caption.
20160804 01:15:45< celmin> Impossible though it sounds.
20160804 01:15:48< vultraz> Not really
20160804 01:15:59< celmin> Yes.
20160804 01:15:59< vultraz> It's still deprecated this series.
20160804 01:16:08< vultraz> Whether it was announced last year or now
20160804 01:16:10< vultraz> It doesn't matter
20160804 01:16:14< vultraz> it's still 1.13
20160804 01:16:33-!- gfgtdf [~chatzilla@x4e368e00.dyn.telefonica.de] has quit [Ping timeout: 240 seconds]
20160804 01:16:38< celmin> The caption issue was probably introduced when I ported [message] to Lua.
20160804 01:16:44-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160804 01:16:46-!- gfgtdf_ is now known as gfgtdf
20160804 01:16:49< vultraz> caption="" makes sense to be blank
20160804 01:16:54< celmin> I know.
20160804 01:17:00< vultraz> caption= null doesn't
20160804 01:17:05< mattsc> vultraz: that’s not the point
20160804 01:17:06< vultraz> (where null = ())
20160804 01:17:21< mattsc> if it is set to (), the caption also does not appear
20160804 01:17:26< vultraz> yes
20160804 01:17:28< mattsc> but that’s also besides the point
20160804 01:17:32< vultraz> that should be fixed
20160804 01:17:39< vultraz> somehow
20160804 01:17:46< celmin> vultraz: That's not true. A key is only null if it's not present at all.
20160804 01:18:23< vultraz> well, blank
20160804 01:18:40< celmin> If it's blank, I'm pretty sure the parser would consider that to be an empty string.
20160804 01:20:17< mattsc> My only real point is that you cannot _silently_ change behavior on people
20160804 01:20:48< mattsc> Whether I personally agree with the decision or not is not important — so I won’t even tell you if I do.  :P
20160804 01:26:16< celmin> …I wonder if there'd be a benefit to implementing some of those utility unit macros as additiona WML tags.
20160804 01:26:22< celmin> Like [loyal_unit] for example.
20160804 01:28:15< vultraz> No
20160804 01:28:45< vultraz> would be worth making it so loyal units didn't need that extra overlay wml tho
20160804 01:29:44< celmin> So that "upkeep=loyal" automatically added the overlay?
20160804 01:29:54< celmin> Hmm, I wonder if that might cause problems though.
20160804 01:30:12< celmin> (Or I think there's an [effect] for overlays, so it could be integrated into the trait?)
20160804 01:30:16< celmin> (But would that cause problems?)
20160804 01:30:31< celmin> Anyway, you seemed pretty quick to say "no" to that.
20160804 01:30:54< celmin> Isn't it a good thing to reduce macro usage?
20160804 01:31:19< vultraz> yes
20160804 01:31:32< vultraz> but it's not worth a new tag to replace one key
20160804 01:31:46< celmin> It's not one key.
20160804 01:31:51< celmin> It's a key and a tag.
20160804 01:32:06< celmin> Or something like that.
20160804 01:32:14< vultraz> the need for the tag should be eliminated 
20160804 01:32:23< vultraz> if you're talking of overlays
20160804 01:32:30< celmin> I was thinking of the trait.
20160804 01:32:38< celmin> Normally loyal units have the loyal trair.
20160804 01:33:03< gfgtdf> celmin: i think there are cases where you don't want the loyal overlay on a loyal unit, like for example, sometime you rather want t a hero overlay or specailly for enemy unity you might want to emphasize the fact that they dont cost upkeep.
20160804 01:33:19< celmin> Though I guess if the overlay was integrated into the trait, and with the new way of specifying traits, it would end up being "trait=loyal"
20160804 01:33:25< celmin> ^traits=loyal
20160804 01:33:49< celmin> So with that I suppose it's not worth it.
20160804 01:33:55< gfgtdf> celmin: i do hoever think tht it does make se sense to change most code to rather add the loyal overlay via an effect instead of direct unit mopdification, maybe make it an optional flag of the loyal effect
20160804 01:34:02< gfgtdf> however*
20160804 01:34:42< celmin> I was thinking of adding a second effect to the loyal trait.
20160804 01:35:20< celmin> [effect]apply_to,add=overlay,loyal.png
20160804 01:35:22< celmin> Or whatever
20160804 01:35:46< mattsc> vultraz: oh, and btw, what I am going to do in my campaigns is just define a macro MESSAGE_CN or something and then do a search and replace  :P
20160804 01:35:51< gfgtdf> celmin: yes sounds like a good idea to me.
20160804 01:36:14< celmin> mattsc: I suppose that works.
20160804 01:37:08< mattsc> celmin: yeah; because, seriously, whether the macro makes sense or not, I don’t see the point of changing thing in that many places …
20160804 01:37:30< celmin> Do you ever actually use a caption?
20160804 01:37:33< mattsc> s/thing/it
20160804 01:37:41< mattsc> yes
20160804 01:38:26< mattsc> most of the time I don’t, but sometimes I do (most often for narrator messages)
20160804 01:38:32< celmin> Wait, I guess {MESSAGE_CN} sounds like it's for the ones that do need captions...
20160804 01:38:50< mattsc> CN = Campaign_Name
20160804 01:38:53< celmin> Oh.
20160804 01:38:56< mattsc> :D
20160804 01:39:50< mattsc> Anyway, it’s not a big deal one way or the other, I just am way too lazy to go through that much effort for something that doesn’t change anything
20160804 01:40:15< mattsc> I’ll probably change it in the small campaigns where it’s only used a couple dozen times.
20160804 02:13:01-!- SigurdFD [~SigurdFD@dynamic-acs-72-23-110-58.zoominternet.net] has joined #wesnoth-dev
20160804 02:25:09-!- gfgtdf [~chatzilla@x4e369370.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 48.0/20160726073904]]
20160804 02:49:08< shadowm> loonycyborg: Did you deploy the fix on the live server?
20160804 02:57:23< shadowm> I guess not.
20160804 03:05:07< shadowm> loonycyborg: So, um, I assume you realize that this kind of stuff is actually very relevant and should be included in the changelog.
20160804 03:05:50< shadowm> And since you are one of the release people I'm pretty sure it'd be very stupid of me to tell you to add something to the changelog, since, well, it's already a thing you work with.
20160804 03:08:39-!- Bonobo [~Bonobo@2001:44b8:254:3200:39d4:5ef:9c96:b5e] has joined #wesnoth-dev
20160804 03:16:29-!- JyrkiVesterinen [~JyrkiVest@87-100-129-99.bb.dnainternet.fi] has joined #wesnoth-dev
20160804 03:24:37-!- celmin [~celticmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The minstrel departs, to spread the music to the masses!]
20160804 03:29:08< shadowm> Okay, fix deployed.
20160804 03:36:47-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160804 03:45:48< celticminstrel> Okay, so I got ancestral's build to work, but at the cost of including two additional libraries.
20160804 03:47:29< celticminstrel> http://celmin.pwcsite.com/wesnoth/Wesnoth-fixedlibs.zip
20160804 03:47:40< celticminstrel> That excludes the Boost fixes.
20160804 03:51:50< celticminstrel> Oh, whoops, missed one in that package. I'll fix it in a moment at the same link. (Not like ancestral's actually around right now anyway.)
20160804 03:59:40< celticminstrel> Okay, I think that should be all of them, though I should probably double-check ancestral's package after he's substituted them...
20160804 04:00:37< celticminstrel> Also, weirdly, wesnothd is referencing /usr/lib/libz and /usr/lib/libbz even though those libraries are packaged with it.
20160804 04:00:59< celticminstrel> Actually, so is the main wesnoth executable.+
20160804 04:02:05< celticminstrel> I guess the debug_ai issue will have to wait until tomorrow.
20160804 04:02:42-!- molt [~molt@37.122.170.190] has quit [Quit: Leaving]
20160804 04:24:37-!- irker340 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160804 04:44:53-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160804 04:46:46-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160804 04:50:07-!- Kwandulin [~Miranda@p200300760F60626F1478B7046D24B45D.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160804 04:51:20-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 258 seconds]
20160804 04:59:01-!- ChipmunkV [~vova@d0017-2-88-172-31-68.fbx.proxad.net] has joined #wesnoth-dev
20160804 04:59:48-!- SigurdFD [~SigurdFD@dynamic-acs-72-23-110-58.zoominternet.net] has quit []
20160804 05:16:58-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 265 seconds]
20160804 05:21:18-!- mjs-de [~mjs-de@x4e3020e4.dyn.telefonica.de] has joined #wesnoth-dev
20160804 05:23:27-!- ChipmunkV [~vova@d0017-2-88-172-31-68.fbx.proxad.net] has quit [Quit: ChipmunkV]
20160804 05:27:02-!- mjs-de [~mjs-de@x4e3020e4.dyn.telefonica.de] has quit [Remote host closed the connection]
20160804 05:27:25-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Read error: Connection reset by peer]
20160804 05:52:27-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160804 05:52:28-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Read error: Connection reset by peer]
20160804 05:56:03-!- TC02 [~quassel@venus.arosser.com] has quit [Ping timeout: 240 seconds]
20160804 06:02:31-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
20160804 06:04:21-!- TC02 [~quassel@venus.arosser.com] has joined #wesnoth-dev
20160804 06:13:32-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160804 06:28:38-!- boucman_work [~boucman@2a02-8428-034f-f800-9e32-0c7c-b391-6223.rev.sfr.net] has joined #wesnoth-dev
20160804 06:35:05-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160804 06:39:39-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20160804 06:47:01-!- Nobun [~nobun@host227-22-dynamic.5-87-r.retail.telecomitalia.it] has joined #wesnoth-dev
20160804 06:47:10-!- Nobun [~nobun@host227-22-dynamic.5-87-r.retail.telecomitalia.it] has quit [Client Quit]
20160804 06:52:53-!- horrowind [~Icedove@2a02:810a:83c0:404:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160804 06:53:06-!- atarocch [~atarocch@194.127.9.104] has joined #wesnoth-dev
20160804 07:11:27-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160804 07:15:55-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20160804 08:19:09-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160804 08:23:48-!- Kwandulin [~Miranda@p200300760F60626F1478B7046D24B45D.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160804 08:26:06-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160804 08:31:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 276 seconds]
20160804 08:59:04< loonycyborg> shadowm: cool, ty
20160804 08:59:18< loonycyborg> it was just really late when I made that fix :P
20160804 08:59:56< loonycyborg> I was too sleepy to do those things..
20160804 09:02:28-!- Kwandulin [~Miranda@p200300760F60626F1DE07CD448155D40.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160804 09:27:05< loonycyborg> JyrkiVesterinen: it seems your reinterpret_cast in random_new.cpp triggers  "dereferencing type-punned pointer will break strict-aliasing rules" warning from gcc for me
20160804 09:27:35-!- Kwandulin_2 [~Miranda@p200300760F6062A81DE07CD448155D40.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160804 09:28:00< JyrkiVesterinen> Okay. I'll need to replace it with an union or something.
20160804 09:28:10< loonycyborg> would be cool
20160804 09:29:48-!- Kwandulin [~Miranda@p200300760F60626F1DE07CD448155D40.dip0.t-ipconnect.de] has quit [Ping timeout: 250 seconds]
20160804 09:29:57-!- Duthlet [~Duthlet@dslb-178-005-055-087.178.005.pools.vodafone-ip.de] has joined #wesnoth-dev
20160804 09:30:11< loonycyborg> I used union to handle issue like this in network_asio.?pp
20160804 09:35:59-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160804 09:40:14-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20160804 09:41:24-!- atarocch [~atarocch@194.127.9.104] has quit [Remote host closed the connection]
20160804 09:41:59-!- atarocch [~atarocch@194.127.9.104] has joined #wesnoth-dev
20160804 09:42:56-!- irker438 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160804 09:42:56< irker438> wesnoth: loonycyborg wesnoth:master 059de7266a2a / changelog: Update changelog https://github.com/wesnoth/wesnoth/commit/059de7266a2a2868848e81d4dbbed73380d9d844
20160804 09:51:36< irker438> wesnoth: Jyrki Vesterinen wesnoth:master 4c2338d61618 / src/random_new.cpp: Avoid a GCC warning about breaking strict-aliasing rules https://github.com/wesnoth/wesnoth/commit/4c2338d61618f817735d60e34ec98232ce4fa45f
20160804 09:56:59-!- JyrkiVesterinen [~JyrkiVest@87-100-129-99.bb.dnainternet.fi] has quit [Quit: Going somewhere. I'll return after about three hours.]
20160804 10:02:04-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160804 10:42:27-!- ChipmunkV [~vova@d0017-2-88-172-31-68.fbx.proxad.net] has joined #wesnoth-dev
20160804 10:43:57-!- Kwandulin_2 is now known as Kwandulin
20160804 10:57:37-!- RatArmy [~RatArmy@om126161124111.8.openmobile.ne.jp] has joined #wesnoth-dev
20160804 11:16:15-!- ChipmunkV [~vova@d0017-2-88-172-31-68.fbx.proxad.net] has quit [Quit: ChipmunkV]
20160804 11:24:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160804 11:25:26-!- hk238 [~kvirc@t224.ip7.netikka.fi] has joined #wesnoth-dev
20160804 11:28:39-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 264 seconds]
20160804 11:42:48-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160804 11:57:57-!- JyrkiVesterinen [~JyrkiVest@87-100-192-155.bb.dnainternet.fi] has joined #wesnoth-dev
20160804 12:01:40-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160804 12:08:23-!- Kwandulin [~Miranda@p200300760F6062A81DE07CD448155D40.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160804 12:51:46-!- irker438 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160804 12:54:40-!- Necrosporus [~Necrospor@unaffiliated/necrosporus] has joined #wesnoth-dev
20160804 13:03:57< JyrkiVesterinen> Some progress with Lua debugging. I got Decoda to compile with some local modifications, and it attaches to Wesnoth and detects the scripts the game loads.
20160804 13:04:15< JyrkiVesterinen> It looks like it doesn't hit breakpoints, though. I'll try to get that working.
20160804 13:04:30< JyrkiVesterinen> (Debugging a debugger, fun...)
20160804 13:12:40-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160804 13:19:49-!- Bonobo [~Bonobo@2001:44b8:254:3200:39d4:5ef:9c96:b5e] has quit [Ping timeout: 250 seconds]
20160804 13:20:11-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160804 13:24:35-!- Kwandulin [~Miranda@p200300760F6062A855FF672C00493E3F.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160804 13:27:49< fabi_> JyrkiVesterinen: Which editor or ide are you using for lua?
20160804 13:28:26< JyrkiVesterinen> I haven't made any changes to Lua code yet. I plan to use Notepad++ for that.
20160804 13:28:39< JyrkiVesterinen> For debugging, I'm trying to use Decoda as we speak.
20160804 13:28:42-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Remote host closed the connection]
20160804 13:29:04< fabi_> "Howl" is an editor that aims at Lua and derivates.
20160804 13:35:07< celticminstrel> ...derivates?
20160804 13:35:15< celticminstrel> That makes it sound like calculus.
20160804 13:35:18< fabi_> MoonScript
20160804 13:35:26< celticminstrel> Ah.
20160804 13:35:38-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160804 13:35:48< fabi_> And Terra
20160804 13:36:18< celticminstrel> What a terrible name.
20160804 13:37:41< fabi_> Yeah, c++ sounds much nicer in my ears.
20160804 13:39:09< celticminstrel> At least it's not a generic proper noun.
20160804 13:39:18< celticminstrel> Uh, that's contradictory I guess.
20160804 13:39:32< celticminstrel> At least it's not a generic word that could be used for almost anything.
20160804 13:39:39< fabi_> Lua is a generic proper noun as well.
20160804 13:40:39< celticminstrel> Oh? Give me at least two uses that don't relate to the scripting language then.
20160804 13:42:40< fabi_> It means moon in Portuguese. I guess there are a lot of uses every day. Considering that over 300.000.000 people speak it.
20160804 13:45:05< celticminstrel> Ah.
20160804 13:45:18< celticminstrel> Still not as bad as Terra, though.
20160804 13:45:28< celticminstrel> (On the other hand, it explains why Terra.)
20160804 13:45:36< celticminstrel> (And MoonScript too for that matter.)
20160804 13:45:36< fabi_> yeah
20160804 13:45:40< fabi_> yeah
20160804 13:46:12< fabi_> And why the editor is called "Howl" as well.
20160804 13:46:32< celticminstrel> Heh.
20160804 13:47:52-!- RatArmy [~RatArmy@om126161124111.8.openmobile.ne.jp] has quit [Ping timeout: 240 seconds]
20160804 13:48:57< JyrkiVesterinen> Update: breakpoints actually worked all along. :) I just placed the first one into a function that wasn't called when I expected.
20160804 13:48:58< JyrkiVesterinen> https://s19.postimg.org/epqth002r/Decoda.png
20160804 13:49:52< celticminstrel> Does it auto-detect the list on the left?
20160804 13:50:05< JyrkiVesterinen> Yes, it does.
20160804 13:50:15< celticminstrel> What's the weird things on the call stack?
20160804 13:50:23< celticminstrel> I guess it doesn't matter, just curious.
20160804 13:51:40< fabi_> JyrkiVesterinen: How is Decoda used? Can I also use it with pure Lua scripts or does it need a host?
20160804 13:51:52-!- RatArmy [~RatArmy@om126161124111.8.openmobile.ne.jp] has joined #wesnoth-dev
20160804 13:52:00< JyrkiVesterinen> celticminstrel: Three of them are internal functions in Windows DLLs. Two are native code functions for which Decoda couldn't figure out the function name.
20160804 13:52:19< celticminstrel> It's unable to load pdb files or something?
20160804 13:52:24< JyrkiVesterinen> fabi_: I think it needs a host, I have never used it with pure scripts.
20160804 13:52:56< JyrkiVesterinen> celticminstrel: Most likely the functions simply aren't mentioned in public PDB files. Likely more Windows internal functions.
20160804 13:53:01< celticminstrel> Ah
20160804 13:53:29< fabi_> https://studio.zerobrane.com/
20160804 13:53:40< celticminstrel> I kinda would've expected to see a bunch of Wesnoth functions on the stack.
20160804 13:54:09< JyrkiVesterinen> fabi_: Decoda is used by specifying the path of the process to launch, command-line parameters and the working directory.
20160804 13:54:18< JyrkiVesterinen> It works by injecting a DLL into the process.
20160804 13:54:30< fabi_> Unlike Decoda, it runs on Windows, Mac OSX, and Linux and supports on-device debugging.
20160804 13:54:38< celticminstrel> Hmmm... do we have a way to load all Lua files in a directory?
20160804 13:54:43< JyrkiVesterinen> That DLL uses Lua's debugging API to implement features like breakpoints.
20160804 13:54:46< fabi_> Sorry, forget the quotation marks.
20160804 13:55:03< fabi_> JyrkiVesterinen: Does that mean Decoda is Windows only?
20160804 13:55:08< JyrkiVesterinen> celticminstrel: There are Wesnoth C++ functions lower down.
20160804 13:55:12< JyrkiVesterinen> fabi_: yes.
20160804 13:55:15< fabi_> oh
20160804 13:55:18< celticminstrel> I imagine injecting DLLs can work on other platforms though.
20160804 13:55:41< celticminstrel> (Apart from not being called DLLs. Also probably requires elevated permissions or something, as it's the sort of thing dangerous programs might do.)
20160804 13:56:20< JyrkiVesterinen> Yes, I believe it's possible. However, doing things like locating the Lua functions in the process requires OS-specific code, and Decoda only has that code for Windows.
20160804 13:56:24< celticminstrel> I'm not exactly sure how gdb/lldb work, but pretty sure no DLL injection is involved.
20160804 13:56:52< celticminstrel> All it needs is to be able to read the process's own memory and send signals to it.
20160804 13:56:59< celticminstrel> I think.
20160804 13:57:28< celticminstrel> Anyway, end of tangent.
20160804 13:57:52< celticminstrel> It would be nice if I don't have to manually enumerate every Lua file in the folder...
20160804 13:58:11< celticminstrel> I could just add a "wesnoth.dofiles" or something...
20160804 14:01:51-!- irker679 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160804 14:01:51< irker679> wesnoth: gfgtdf wesnoth:master 531b91359a0f / src/units/unit.cpp: fix wrong underlying_id parsing on 64 bit systems. https://github.com/wesnoth/wesnoth/commit/531b91359a0fb465883839bffbe51f29e35d56c7
20160804 14:02:14< celticminstrel> 64-bit systems, or 64-bit builds... >_>
20160804 14:20:01< celticminstrel> Hmm, I can only see four WML tags that were definitely not properly ported, so that's nice I guess.
20160804 14:20:19< celticminstrel> Three I'm not sure of, as well, so as many as seven.
20160804 14:24:55< fabi_> celticminstrel: Not properly ported WML tags?
20160804 14:25:28< celticminstrel> fabi_: Cases where someone basically cut-and-pasted the WML tag code into a Lua API function, and made the Lua implementation simply call that.
20160804 14:25:56< celticminstrel> I think the shroud tag-pair might also be a culprit actually... though a milder case, perhaps...
20160804 14:27:57< celticminstrel> Hmm, no, they're definitely a culprit.
20160804 14:28:11< celticminstrel> And [scroll] too. So that means 7-10 now.
20160804 14:29:13< irker679> wesnoth: Jyrki Vesterinen wesnoth:master 3cbf249b663f / src/play_controller.cpp: play_controller::check_victory(): use underscore in event name https://github.com/wesnoth/wesnoth/commit/3cbf249b663ff957d936ac1612bfe848dcea98dd
20160804 14:30:15< celticminstrel> I want to revert the "local victory" event. It's a ridiculous idea.
20160804 14:30:26< celticminstrel> (Not going to unless proper synching can be attained, but still.)
20160804 14:30:28-!- horrowind [~Icedove@2a02:810a:83c0:404:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160804 14:34:07-!- gfgtdf [~chatzilla@x4e369370.dyn.telefonica.de] has joined #wesnoth-dev
20160804 14:34:13-!- RatArmy [~RatArmy@om126161124111.8.openmobile.ne.jp] has quit [Ping timeout: 244 seconds]
20160804 14:36:42< gfgtdf> JyrkiVesterinen: are you 'jyrkive' in github? If yes, please change eigher _all_ event (side turn, pre advance  ...) to use underscrores or revert 3cbf249b6, hving one event name use undercoed and all other use spaces i the c++ code doesnt reall make finding things easier
20160804 14:37:19< celticminstrel> The answer to your question should be incredibly obvious.
20160804 14:38:04< JyrkiVesterinen> All other? Has it been the standard to use spaces in event names when firing them from C++?
20160804 14:39:05< JyrkiVesterinen> I don't see what's the point to consistently use spaces at call sites and then convert them to underscores with code...
20160804 14:39:18< JyrkiVesterinen> I'll convert the other events as well, then.
20160804 14:39:35< celticminstrel> Yeah, then half the conversion could be removed.
20160804 14:39:49< celticminstrel> (Of course, the event names read from WML would still need to be converted.)
20160804 14:40:14< vultraz> ^
20160804 14:40:30< celticminstrel> Why are you pointing at my post?
20160804 14:40:37< vultraz> I agree
20160804 14:40:57< vultraz> "^" means "this, this is good"
20160804 14:41:06< gfgtdf> JyrkiVesterinen: well i dont care that much whether its spaces or undersctores in the c++ code i just think we shoudl make them all consitent to make it easier to find them.
20160804 14:41:24< celticminstrel> Agreed.
20160804 14:41:57< JyrkiVesterinen> My position is that they're easier to find (we avoid a "gotcha") if the event name is the same in firing code and handling code.
20160804 14:42:08< celticminstrel> (Does anyone else agree that the "local victory" event was a bad idea?)
20160804 14:43:19< JyrkiVesterinen> Not suprising: when I noticed that the Lua code had been passed an enemies_defeated event, I scanned the codebase for that text. I was surprised that the event didn't seem to be sent from anywhere.
20160804 14:43:44< JyrkiVesterinen> I figured that the event name is probably generated dynamically. And it kind of was.
20160804 14:48:07-!- Kwandulin [~Miranda@p200300760F6062A855FF672C00493E3F.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160804 14:52:46-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20160804 14:53:01-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Remote host closed the connection]
20160804 14:54:06-!- vultraz_iOS [uid24821@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160804 14:55:19< gfgtdf> celticminstrel: np, what is is that you don't like about it ?
20160804 14:55:23< gfgtdf> no*
20160804 14:55:36< celticminstrel> Everything.
20160804 14:56:32< celticminstrel> It's like, a hack rather than a proper solution.
20160804 14:56:45< celticminstrel> (Anyway, I was asking everyone except you with that question.)
20160804 14:56:47 * fabi_ enjoys local victories more than remote ones.
20160804 14:57:10< gfgtdf> celticminstrel: i think its a very peoper solution. And its not like you have to use that event.
20160804 14:57:30< celticminstrel> "you don't have to use it" is not a good argument in favour of a bad feature.
20160804 14:57:46< celticminstrel> It's not a proper solution. A proper solution is synchronization.
20160804 14:58:29< celticminstrel> (Admittedly, synchronization is hard in this case. But that's no reason to hack in something like this.)
20160804 14:58:41< fabi_> You boys should not introduce stuff that consolidates the ugly fat client solution anymore.
20160804 14:59:09< celticminstrel> XD I got a Lua stack overflow.
20160804 15:00:19< celticminstrel> Because I overrode wesnoth.modify_side to call wesnoth.wml_actions.modify_side and forgot to save my changes to wesnoth.wml_actions.modify_side to make it not call wesnoth.modify_side.
20160804 15:01:31< celticminstrel> I wonder if it's possible to catch a stack overflow.
20160804 15:01:44< celticminstrel> (A native one, not a Lua one.)
20160804 15:01:51< celticminstrel> Probably not a good idea even if it is.
20160804 15:01:57< celticminstrel> (Unless you're writing a debugger.)
20160804 15:02:55< celticminstrel> Yay, the side userdata is mutable. That saves me some work.
20160804 15:03:40< celticminstrel> I wonder if there's a Lua builtin to sort an array-like table and remove duplicates...
20160804 15:03:54< celticminstrel> Actually wait, I don't need it.
20160804 15:04:07< celticminstrel> [modify_side]recruit= replaces rather than augments the recruit list.
20160804 15:04:28< celticminstrel> ...does gmatch return an iterator or a table?
20160804 15:05:19< JyrkiVesterinen> It returns an iterator if I recall correctly.
20160804 15:05:43< celticminstrel> I don't suppose there's a way to turn that into a table without an explicit loop.
20160804 15:06:25< celticminstrel> gfgtdf: Wasn't there something about controller changes not working in some situations?
20160804 15:12:07< fabi_> celticminstrel: The WML wiki knows not about "local victory".
20160804 15:12:29-!- oldlaptop [~quassel@50.36.241.195] has quit [Ping timeout: 250 seconds]
20160804 15:12:32< celticminstrel> gfgtdf: So uhh, I tried changing controller from ai to human, and the game crashed. This is absolutely not good.
20160804 15:13:16< celticminstrel> gfgtdf: Don't use assert to guarantee that the context is synced. Use an if statement and print a warning or something if it's not. Seriously, assert is the worst choice for something like that.
20160804 15:13:45< celticminstrel> (I'm going to make that change myself in this case since it's easy.)
20160804 15:14:02< celticminstrel> (And if it still crashes, I'll print an error instead and do nothing.)
20160804 15:14:27< celticminstrel> (Wait, hang on, why is it calling...)
20160804 15:15:05 * celticminstrel is still going to change that to an if statement, but... investigating a bit more closely first.
20160804 15:15:36-!- oldlaptop [~quassel@50.36.241.195] has joined #wesnoth-dev
20160804 15:16:08-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160804 15:18:28-!- JyrkiVesterinen [~JyrkiVest@87-100-192-155.bb.dnainternet.fi] has quit [Quit: Going somewhere]
20160804 15:21:18 * celticminstrel notes that it's entirely possible that gfgtdf was not to blame for this particular assert.
20160804 15:21:45< celticminstrel> ...I wonder how much of a task it would be to completely purge asserts from the codebase.
20160804 15:22:34< celticminstrel> Though there are probably some legitimate uses of them.
20160804 15:22:59< fabi_> celticminstrel: Asserts are useful. The usual used solution is to compile the production binary with asserts disabled.
20160804 15:23:16< celticminstrel> The problem is that asserts are widely misused in Wesnoth.
20160804 15:23:32< fabi_> like exceptions
20160804 15:23:35< celticminstrel> They should be used only for conditions that you're pretty sure are going to be true.
20160804 15:23:55< gfgtdf> celticminstrel: where exactly did you try to change the controllers ?
20160804 15:24:03< celticminstrel> Yeah sure, but the misuse of asserts is separate from and (probably) unrelated to the misuse of exceptions.
20160804 15:24:05< celticminstrel> gfgtdf: Lua console.
20160804 15:24:09< gfgtdf> celticminstrel: you have a stacktrace about that assertion failure ?
20160804 15:24:22< celticminstrel> I can get you one if you really want?
20160804 15:24:29< celticminstrel> Just let me trigger it again.
20160804 15:24:32< gfgtdf> celticminstrel: y it would be useful. thx
20160804 15:25:01< celticminstrel> Obviously this is an unsafe place to change controllers and would cause OOS if it were done in an MP game, but even so, it shouldn't crash.
20160804 15:25:09< fabi_> Asserts should be used to guard the invariants of objects.
20160804 15:25:26< celticminstrel> They shouldn't be used for things that you're not sure will be true.
20160804 15:25:58< celticminstrel> (This applies to C asserts. Asserts in other languages aren't as bad.)
20160804 15:26:13< celticminstrel> (Most of them that I know of actually translate to an exception.)
20160804 15:32:56-!- ChipmunkV [~vova@d0017-2-88-172-31-68.fbx.proxad.net] has joined #wesnoth-dev
20160804 15:34:34< gfgtdf> celticminstrel: you have teh stacktrace?
20160804 15:35:13< celticminstrel> Yeah, hang on.
20160804 15:37:55< celticminstrel> ...so, Firefox is not responding. Fun.
20160804 15:43:22-!- atarocch [~atarocch@194.127.9.104] has quit [Ping timeout: 265 seconds]
20160804 15:46:08< celticminstrel> gfgtdf: http://pastebin.com/81PaezE1
20160804 15:52:57-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160804 15:52:58-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160804 15:53:04-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160804 15:53:54< irker679> wesnoth: gfgtdf wesnoth:master e75325a39959 / src/synced_context.cpp: fix assertion in synced_context::ask_server_choice https://github.com/wesnoth/wesnoth/commit/e75325a399593d0c9e50a0b41b36d290ee4f6db9
20160804 15:53:58< gfgtdf> celticminstrel: ^
20160804 15:54:45-!- JyrkiVesterinen [~JyrkiVest@87-100-192-155.bb.dnainternet.fi] has joined #wesnoth-dev
20160804 15:55:15< celticminstrel> What the heck? After testing wesnoth.sides[1].shroud, I'm now getting ilua strictness errors when accessing wesnoth.sides.
20160804 15:55:38< celticminstrel> Ohhh. I think I accidentally a comma.
20160804 15:55:51< celticminstrel> Doesn't help that I can't tell the two apart in this font. ><
20160804 15:59:48< zookeeper> asserts are weird. on one hand i agree that asserts should test for things you think would never become true, but if i try to do that, how do i decide which asserts to add? there's an infinite number of conditions that i think would be impossible to become true but which would be really bad if they did.
20160804 16:02:42< zookeeper> ok, in practise it's often clear what you might want to test for (function arguments, etc), but still.
20160804 16:06:52< JyrkiVesterinen> zookeeper: One example is that lately (for Monte Carlo simulation) I needed to scale the values in a std::vector<double> so that their sum is 1. (Two or four of them, to be exact.)
20160804 16:07:18< JyrkiVesterinen> My code already gets the needed multiplier from calling code, so it doesn't need to loop through the vector to calculate it.
20160804 16:07:37< JyrkiVesterinen> But I added an assertion that verifies in debug builds that the sum is indeed 1 after scaling.
20160804 16:08:38< JyrkiVesterinen> The assertion actually uncovered a bug. (An off-by-one error elsewhere, making the simulation code slightly overestimate the probability to not be hit.)
20160804 16:09:01< JyrkiVesterinen> https://github.com/wesnoth/wesnoth/blob/master/src/attack_prediction.cpp#L1462
20160804 16:09:52< zookeeper> sure, seems like a good idea there
20160804 16:22:57< celticminstrel> Why is the flag an empty string?
20160804 16:25:52< celticminstrel> Ah, I see.
20160804 16:33:55-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160804 16:39:46-!- atarocch [~atarocch@ipbcc2d6b3.dynamic.kabel-deutschland.de] has joined #wesnoth-dev
20160804 16:39:52-!- gfgtdf [~chatzilla@x4e369370.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 48.0/20160726073904]]
20160804 16:48:27-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160804 17:07:34-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Remote host closed the connection]
20160804 17:18:38-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160804 17:19:29-!- Kwandulin [~Miranda@p200300760F6062A83D15D38628549D4B.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160804 17:29:40-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160804 17:32:25-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 252 seconds]
20160804 17:32:25-!- wedge010 is now known as wedge009
20160804 17:33:36< celticminstrel> Hmm, is there currently a way to get the readonly context...
20160804 17:36:16< celticminstrel> Seems like the answer is probably yes?
20160804 17:37:51< celticminstrel> I don't really get all this indirection stuff though.
20160804 17:38:01< celticminstrel> It seems like ai_composite is itself a readonly context.
20160804 17:38:13< celticminstrel> So why is there a separate pointer to a separate readonly context?
20160804 17:38:41< celticminstrel> Which is indirectly used to construct the ai_composite...
20160804 17:47:02< celticminstrel> Anyway, I've almost finished properly porting [modify_side] to Lua. New API functions are wesnoth.switch_ai and wesnoth.append_ai, and it lets [modify_side][ai] add goals and stages as well as facets. Also a number of side proxy keys are added or made mutable.
20160804 17:47:27< celticminstrel> Kinda don't want to test it right now, so I'll probably commit it to a branch and test later.
20160804 17:48:32-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160804 17:50:17< celticminstrel> I should check if wesnoth.modify_side is documented. If it isn't, there's probably no need to announce its deprecation. It couldn't hurt to check if addons use it, but I suspect they would've been calling wesnoth.wml_actions.modify_side instead.
20160804 17:56:10< celticminstrel> gfftdf: Do you think it's a good idea for wesnoth.sides[1].flag to return game_config::images::flag instead of an empty string?
20160804 17:57:13< celticminstrel> I guess the alternative would be to let Lua access the game_config value directly (though I imagine it already can).
20160804 18:04:58-!- horrowind [~Icedove@2a02:810a:83c0:404:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160804 18:07:12< irker679> wesnoth: Charles Dang wesnoth:master 95fecb9d909d / data/gui/widget/unit_preview_pane.cfg: Small code and border cleanups https://github.com/wesnoth/wesnoth/commit/95fecb9d909dec555ab07beeb44d7f260c2acd78
20160804 18:13:36-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 276 seconds]
20160804 18:13:54< celticminstrel> Okay, I'm going to merge iceiceice's PR with some compatibility tweaks.
20160804 18:14:20< celticminstrel> Unless someone objects.
20160804 18:14:25< vultraz> fix the indent too
20160804 18:14:32< celticminstrel> Huh?
20160804 18:14:50< vultraz> Are you referring to 720?
20160804 18:14:54< celticminstrel> Yes.
20160804 18:15:11< vultraz> It has some bad indent.
20160804 18:15:14< vultraz> Last i checked.
20160804 18:15:17< celticminstrel> I'm not seeing it.
20160804 18:15:24< celticminstrel> Check again maybe?
20160804 18:15:51< vultraz> Line 51 in the pr
20160804 18:16:10< celticminstrel> What file?
20160804 18:16:19< vultraz> formatter
20160804 18:18:02-!- hk238 [~kvirc@t224.ip7.netikka.fi] has quit [Quit: http://www.kvirc.net/ 4.9.1 Aria]
20160804 18:18:37-!- vultraz_iOS [uid24821@wesnoth/developer/vultraz] has quit [Quit: Connection closed for inactivity]
20160804 18:19:11< celticminstrel> Oh, I see it now.
20160804 18:19:25< irker679> wesnoth: Jyrki Vesterinen wesnoth:master 9336b767646e / src/menu_events.cpp: Fix scenario end events not being run on debug-killing the enemy leader https://github.com/wesnoth/wesnoth/commit/9336b767646eb08243d88672787d12709ec29014
20160804 18:19:54< vultraz> ah, nice :D
20160804 18:19:56< JyrkiVesterinen> ^ Way to go. Hunt a bug for two days and fix it by removing one character. ;)
20160804 18:20:15< celticminstrel> What XD
20160804 18:20:33< vultraz> #JustWesnothThings
20160804 18:21:10< vultraz> I'll mark the bug as fixed
20160804 18:21:52< JyrkiVesterinen> Was there a bug report about it?
20160804 18:22:22< celticminstrel> How long has it been since the release was tagged?
20160804 18:22:49< vultraz> at least a week, I think
20160804 18:22:56< vultraz> where is that macOS package!
20160804 18:23:06< vultraz> JyrkiVesterinen: https://gna.org/bugs/index.php?24448
20160804 18:23:41< celticminstrel> If he doesn't show up today, you should probably post the announcement...
20160804 18:23:57< celticminstrel> I actually have a working one, but I can't upload it.
20160804 18:24:08< JyrkiVesterinen> vultraz: I see, thanks. I went to hunt for the bug because I hit it myself.
20160804 18:24:10< celticminstrel> (Also I'm not certain it works on 1.10 etc)
20160804 18:24:53< vultraz> it's almost 6 am for me
20160804 18:24:59< vultraz> ima catch some sleep
20160804 18:25:08< celticminstrel> Okay, then post it tomorrow I guess?
20160804 18:25:12< vultraz> I'll announce within 12 hours
20160804 18:25:20< celticminstrel> Hopefully he'll show up by then.
20160804 18:26:14-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160804 18:31:16< celticminstrel> If possible, please don't commit anything until I push this. Should only be a couple of minutes.
20160804 18:31:39< vultraz> no problem
20160804 18:35:29< celticminstrel> Only 64 source files to rebuild, huh.
20160804 18:35:46< celticminstrel> I expect it to error for me.
20160804 18:35:57< vultraz> btw
20160804 18:36:03-!- gfgtdf [~chatzilla@x4e369370.dyn.telefonica.de] has joined #wesnoth-dev
20160804 18:36:06< vultraz> could you sometime go through the FRs on the tracker
20160804 18:36:15< celticminstrel> Uh. Not sure, maybe?
20160804 18:36:18< vultraz> specifically those related to WML
20160804 18:36:26< vultraz> and either implement the requests
20160804 18:36:28< vultraz> or close them
20160804 18:36:28< celticminstrel> I generally don't touch the tracker unless explicitly pointer at an issue.
20160804 18:36:34< vultraz> at your discretion 
20160804 18:36:46< celticminstrel> But I guess I could do that.
20160804 18:37:04< celticminstrel> ^pointed
20160804 18:37:17< vultraz> for example this https://gna.org/bugs/index.php?24576
20160804 18:37:23< vultraz> (color= in [message])
20160804 18:37:31-!- mjs-de [~mjs-de@x4db60cb9.dyn.telefonica.de] has joined #wesnoth-dev
20160804 18:37:34< vultraz> worth it? dunno. if you don't think so, close it
20160804 18:37:34< celticminstrel> Didn't Pentarctagon implement that already?
20160804 18:37:54< celticminstrel> I hope it wasn't forgotten in the changelog...
20160804 18:38:28< celticminstrel> Oh huh, the ref-qualifiers actually work for me. o.o
20160804 18:38:55< celticminstrel> Well, still need the MSVC 2013 protection though.
20160804 18:39:09< vultraz> uh
20160804 18:39:20< vultraz> why are there a bunch of formatting options for [message]
20160804 18:39:32< celticminstrel> Because Pango markup is ugliest.
20160804 18:40:09< vultraz> no, i mean we must have forgotten to put it in the changelog
20160804 18:40:13< celticminstrel> Frankly, the old formatting markup was far better than Pango (even though it wasn't very good).
20160804 18:40:17< vultraz> and i guess that fr is implemented
20160804 18:40:19< celticminstrel> Okay, that's not so good.
20160804 18:40:25< celticminstrel> That it's missing from the changelog, I mean.
20160804 18:40:49< celticminstrel> I guess we can add it after the fact, but it still won't be listed in the officially released changelog for that version.
20160804 18:41:50< vultraz> anyway, you can filter on FRs in the WML category 
20160804 18:41:52< vultraz> there are 72
20160804 18:42:33< vultraz> some will certainly be invalid
20160804 18:42:41< vultraz> since they date back to 2005
20160804 18:42:45< vultraz> but who knows
20160804 18:43:10< celticminstrel> I guess _MSC_VER < 1900 should work...
20160804 18:43:59< celticminstrel> Not sure what the number is for MSVC 2013 and need to launch the browser to check.
20160804 18:44:09< vultraz> some of these we probably don't want to do
20160804 18:44:43< vultraz> celticminstrel: 1800
20160804 18:44:58< celticminstrel> I wonder if values between 1800 and 1900 mean anything.
20160804 18:45:13< celticminstrel> I guess  <1900 is probably safest then.
20160804 18:46:52-!- gfgtdf [~chatzilla@x4e369370.dyn.telefonica.de] has quit [Ping timeout: 240 seconds]
20160804 18:47:15< celticminstrel> Oh right, I should check if GCC 4.7 supports ref-qualifiers...
20160804 18:48:05< vultraz> we're moving to 4.8
20160804 18:48:07< vultraz> check 4.8
20160804 18:48:27< celticminstrel> Guessing it probably does by 4.8.
20160804 18:48:38 * vultraz restarts pc
20160804 18:48:44-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Remote host closed the connection]
20160804 18:48:45< celticminstrel> We did already confirm that MSVC 2013 does not, as I recall.
20160804 18:48:51 * celticminstrel poke JyrkiVesterinen 
20160804 18:49:25< JyrkiVesterinen> IIRC, GCC 4.8.1 has full C++11 support, except <regex>.
20160804 18:50:25< JyrkiVesterinen> I haven't tested ref-qualifiers myself, but according to this page MSVC2013 doesn't support them.
20160804 18:50:26< JyrkiVesterinen> http://en.cppreference.com/w/cpp/compiler_support
20160804 18:50:49< celticminstrel> Unfortunately, changes to global.hpp require a full rebuild. Well, I guess that's to be expected.
20160804 18:52:34< celticminstrel> It's kinda odd that it built fine before but now gives compile errors with the preprocessor statements in...
20160804 18:55:29-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160804 18:55:48< celticminstrel> Huh? Why did most of the errors suddenly vanish?
20160804 18:59:31-!- celmin [~celticmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160804 19:03:20< celmin> Well, the build is almost finished, but still with errors.
20160804 19:03:52< celmin> My compiler seems to feel that the &-qualified and &&-qualified versions are ambiguous, which leads me to believe that it recognizes the syntax but doesn't understand it.
20160804 19:07:45-!- edgrey [~edgrey@178.204.120.31] has joined #wesnoth-dev
20160804 19:12:39< celmin> Apple really makes clang version detection difficult.
20160804 19:17:18-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 276 seconds]
20160804 19:36:14< celmin> Okay, redid the feature testing, compiling again.
20160804 19:36:21< celmin> Sorry this took so long.
20160804 19:37:03< celmin> Still not sure it's working, either... :|
20160804 19:37:23< celmin> Maybe I'll push and break the MSVC2013 build and then push the fix a bit later.
20160804 19:37:39< celmin> (Well, the expected fix. I'm not testing on MSVC2013 after all.)
20160804 19:41:56< irker679> wesnoth: gfgtdf wesnoth:master a0233a1d4df6 / src/ (menu_events.cpp synced_commands.cpp): fix debug kill command beeing unsynced https://github.com/wesnoth/wesnoth/commit/a0233a1d4df6463323b06bdf3f32f3b1345cf574
20160804 19:42:12< celmin> Argh.
20160804 19:42:32< celmin> :/
20160804 19:42:48< celmin> Now I have to rebase. With a merge commit. Which is more of a pain than usual.
20160804 19:46:10< celmin> Well, at least it seems like it's working now. Not sure what was wrong before.
20160804 19:47:56-!- gfgtdf [~chatzilla@x4e369370.dyn.telefonica.de] has joined #wesnoth-dev
20160804 19:54:53-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160804 19:55:23< vultraz> well, I now have ubuntu bash
20160804 19:55:25< vultraz> on windows
20160804 20:01:56< irker679> wesnoth: Chris Beck wesnoth:master 33b0ea50b9d5 / src/ (5 files in 2 dirs): give formatter an x-value conversion to std::string, use in some examples https://github.com/wesnoth/wesnoth/commit/33b0ea50b9d5eafc45caea490c3d11e3db0d1c98
20160804 20:01:59< irker679> wesnoth: Celtic Minstrel wesnoth:master 46aef439132a / src/ (5 files in 2 dirs): Merge pull request #720 from cbeck88/formatter https://github.com/wesnoth/wesnoth/commit/46aef439132a1918975ddeffebcba4a70ab18be7
20160804 20:02:00< irker679> wesnoth: Celtic Minstrel wesnoth:master 8689bedde787 / src/formatter.hpp: Fix indentation https://github.com/wesnoth/wesnoth/commit/8689bedde7870d100f1dcbd9b00ef9b84b7ded8d
20160804 20:02:02< irker679> wesnoth: Celtic Minstrel wesnoth:master 6b019cbc2d9f / src/global.hpp: Macros intended for detecting C++11 features https://github.com/wesnoth/wesnoth/commit/6b019cbc2d9f6ace51df562193209525b8bf0d93
20160804 20:02:04< irker679> wesnoth: Celtic Minstrel wesnoth:master e0392bef87f3 / src/formatter.hpp: Use preprocessor checks in formatter.hpp https://github.com/wesnoth/wesnoth/commit/e0392bef87f3efd6fac968e07a76ec759b6f2f36
20160804 20:02:17< celmin> Please test. Especially wedge009 or JyrkiVesterinen.
20160804 20:02:34< celmin> Actually, more JyrkiVesterinen than wedge009. I seem to recall the latter was using 2015.
20160804 20:02:57< JyrkiVesterinen> I'm going to bed soon. I'll test it tomorrow.
20160804 20:03:08< celmin> Sure.
20160804 20:03:23< celmin> I don't think a problem is likely, anyway.
20160804 20:03:58< celmin> BTW, thanks to 6b019cbc everyone will need to do a full rebuild. Sorry!
20160804 20:06:55< vultraz> damn msvc 2013 :(
20160804 20:07:15< celmin> Well, it's because global.hpp, so that won't be necessary for future things of this type.
20160804 20:08:15< vultraz> well, more like I meant the fact that we have to do such things in the first place
20160804 20:08:19-!- JyrkiVesterinen [~JyrkiVest@87-100-192-155.bb.dnainternet.fi] has quit [Quit: Sleep]
20160804 20:08:29< vultraz> but we need 2013 support
20160804 20:09:21< celmin> So with the new macros in global.hpp, we can conditionally use constexpr, noexcept, and inheriting constructors as well.
20160804 20:09:41< celmin> (Actually, I probably should've handled constexpr and noexcept a little differently…)
20160804 20:13:18-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160804 20:13:47< irker679> wesnoth: Celtic Minstrel wesnoth:master 79fc5cb77ca9 / src/global.hpp: Redo constexpr/noexcept macros to function as keywords https://github.com/wesnoth/wesnoth/commit/79fc5cb77ca961f59c9a17749ecbed18e20142bb
20160804 20:14:25< celmin> Might need to add back HAVE_CONSTEXPR later. We'll think about that when we get to it.
20160804 20:18:11-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Client Quit]
20160804 20:18:56-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160804 20:39:00-!- edgrey [~edgrey@178.204.120.31] has quit [Quit: Konversation terminated!]
20160804 20:40:18-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160804 20:41:21-!- lipkab [~the_new_l@host-91-147-210-58.biatv.hu] has joined #wesnoth-dev
20160804 20:41:21< fabi_> celmin: variable in [foreach] and in [for] have a different meaning. I strongly suggest to call the array to iterate over always "array".
20160804 20:41:36-!- Kwandulin [~Miranda@p200300760F6062A83D15D38628549D4B.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160804 20:41:49< celmin> Geh, now I have to look up what you mean because I don't remember.
20160804 20:42:06< tad_> Question on ~data/core/images/units/*.png and 'known incorrect sRGB profile' errors on stdout ...
20160804 20:42:13< celmin> Oh hi.
20160804 20:42:34< fabi_> celmin: https://wiki.wesnoth.org/ConditionalActionsWML#.5Bfor.5D
20160804 20:42:43< celmin> Thanks for the boost.
20160804 20:42:59< celmin> Hmm, I see what you mean.
20160804 20:43:34< celmin> Okay, that's a trivial change, so I'll go ahead and do it.
20160804 20:43:40< celmin> (Or actually, you could do it if you want.)
20160804 20:44:02< tad_> I tested human-magi unit images for known point of failure HttT S03 arch mage. Ran through pngcrush to strip sRGB profile and, yep, error goes away.
20160804 20:44:30< celmin> So basically some of our png images need to have the profiles stripped out.
20160804 20:44:39< celmin> Maybe we can get zookeeper or someone to do that.
20160804 20:44:49< vultraz> unless tad_'s done it already
20160804 20:44:51< celmin> (I wonder if the optipng thing does that.)
20160804 20:45:04< celmin> I suppose, but tad_ does not have commit access.
20160804 20:45:08< vultraz> he can PR it
20160804 20:45:17< vultraz> (is tad_ Gregory?)
20160804 20:45:20< celmin> Yes.
20160804 20:45:21< tad_> Yep. Looks like it. I'll look to see if there is a way to check and only strip the effected files. pngcrush re-compresses and the files typically grew about 5% in size.
20160804 20:45:24< tad_> Yes
20160804 20:45:46 * celmin has been calling him tad in general here, even when he's not around.
20160804 20:46:18< celmin> There's some sort of optipng tool somewhere.
20160804 20:46:26< celmin> I'm not sure if it's in the repo or what.
20160804 20:46:28< vultraz> ive found the cause of the "Unknown image ~RC(>1)" warnings, btw.
20160804 20:46:32< celmin> But maybe it'll do that?
20160804 20:46:41< celmin> Stripping profiles I mean.
20160804 20:46:53< tad_> pngcrush is what I find refered to a lot and it's in arch/community so easy to keep up-to-date.
20160804 20:47:23< tad_> I will look for something I can use to generate a list of effected files. I'd hate to have to PR the entire image library.
20160804 20:47:49< celmin> So fabi_: If you want to change [foreach]variable to [foreach]array, go for it. We could also leave a compatibility check since it has been in two dev releases, I suppose.
20160804 20:47:59< celmin> Though since it's only been dev releases, maybe we don't need that.
20160804 20:48:17< celmin> (If you do it, don't forget to mention it in changelog.)
20160804 20:49:17< fabi_> celmin: That is a matter of c++ coding? I have no c++ development setup right now.
20160804 20:49:25< celmin> No, it's Lua coding.
20160804 20:49:33< celmin> data/lua/wml-flow.lua
20160804 20:49:46< tad_> OH .. and on HttT, I've finished my list of issues and am about to start code-coverage pass so I'm nearing the end of what I want to do with it.
20160804 20:49:51< celmin> Nice.
20160804 20:50:12< fabi_> celmin: Okay, I will have to lift that code to Irdya anyway at some point.
20160804 20:53:23-!- mjs-de [~mjs-de@x4db60cb9.dyn.telefonica.de] has quit [Remote host closed the connection]
20160804 20:55:06-!- travis-ci [~travis-ci@ec2-54-91-37-166.compute-1.amazonaws.com] has joined #wesnoth-dev
20160804 20:55:07< travis-ci> wesnoth/wesnoth#10123 (master - e0392be : Celtic Minstrel): The build has errored.
20160804 20:55:07< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/149859931
20160804 20:55:07-!- travis-ci [~travis-ci@ec2-54-91-37-166.compute-1.amazonaws.com] has left #wesnoth-dev []
20160804 20:55:14< celmin> Fuuuun!
20160804 20:55:25< celmin> Well, errored usually means it's Travis's fault though.
20160804 20:59:18< gfgtdf> celmin: does the 1.13.5 macosx build work now ?
20160804 20:59:41< celmin> The one that ancestral posted works with my substituted libraries and the fixed Boost libraries.
20160804 20:59:55< celmin> So I guess that's a yes.
20160804 21:00:18< celmin> It would be best if we could verify that ancestral could actually build with my libraries as well.
20160804 21:00:25< celmin> But it's more important that it runs, I guess.
20160804 21:00:40-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160804 21:05:27-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160804 21:10:23-!- horrowind [~Icedove@2a02:810a:83c0:404:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160804 21:14:16< gfgtdf> vultraz: you can do the 1.13.5 release then ?
20160804 21:14:35< celmin> I think he's going to post the announcement without the Mac build.
20160804 21:14:45< celmin> (Since it needs to be on SourceForge first.)
20160804 21:15:23< gfgtdf> celmin: hmm ok, iirc 1.12.6 was also announced before the mac build was avaiable
20160804 21:15:37< gfgtdf> celmin: it can still be added later
20160804 21:15:46< celmin> I'm guessing it won't be much longer.
20160804 21:15:57< celmin> But it basically depends on ancestral finding the time.
20160804 21:17:19< zookeeper> tad, celmin, AFAIK zopflipng is the best compressor
20160804 21:17:36-!- lipkab [~the_new_l@host-91-147-210-58.biatv.hu] has quit [Quit: Leaving]
20160804 21:17:38< zookeeper> although not sure if it can/will strip chunks like that
20160804 21:19:01-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 252 seconds]
20160804 21:24:19-!- vultraz [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20160804 21:24:19-!- vultraz [~chatzilla@124.109.10.167] has quit [Changing host]
20160804 21:24:19-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160804 21:27:48-!- boucman_work [~boucman@2a02-8428-034f-f800-9e32-0c7c-b391-6223.rev.sfr.net] has quit [Quit: Leaving.]
20160804 21:28:11< vultraz> celmin: later
20160804 21:28:14< celmin> ?
20160804 21:29:02< vultraz> I will release later
20160804 21:29:11< vultraz> well
20160804 21:29:15< vultraz> announce the release
20160804 21:31:04-!- RatArmy [~RatArmy@om126161124111.8.openmobile.ne.jp] has joined #wesnoth-dev
20160804 21:32:36-!- Appleman1234 [~Appleman1@KD036012031015.au-net.ne.jp] has quit [Ping timeout: 250 seconds]
20160804 21:35:00< crimson_penguin> vultraz: wait, you're the Mac packager now? :O
20160804 21:35:13< celmin> Note his correction.
20160804 21:35:27< crimson_penguin> oh, the release announcer
20160804 21:35:44< crimson_penguin> you know, having been the Mac packager before may have helped me get this job; one of my interviewers had played my Mac packages of Wesnoth :D
20160804 21:36:00< celmin> Huh...
20160804 21:36:07-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160804 21:37:23< vultraz> crimson_penguin: ancestral is the packager
20160804 21:37:31< crimson_penguin> ah, cool
20160804 21:37:51< vultraz> crimson_penguin: I don't package anything, actually, I just manage things as the community manager and project admin.
20160804 21:38:01< crimson_penguin> well that's cool
20160804 21:40:53< vultraz> you're in NY right now, correct?
20160804 21:41:34< crimson_penguin> I am
20160804 21:41:57< vultraz> ah, I remembered correctly 
20160804 21:42:12< crimson_penguin> live in Brooklyn, work in Manhattan
20160804 21:42:36< vultraz> isn't rent in Brooklyn still obscene?
20160804 21:42:57< crimson_penguin> compared to Vancouver, yeah, but compared to Manhattan it's a bit better
20160804 21:43:11< crimson_penguin> the pay makes up for it though
20160804 21:43:19< vultraz> Oh?
20160804 21:43:56< vultraz> So you say being the mac packager for wesnoth might have helped land you a well-paying job that makes up for the high cost of living in Brooklyn.
20160804 21:44:02< vultraz> Iiinnterestting.
20160804 21:46:41< crimson_penguin> I mean, I don't think it was really a deciding factor, but it may have helped. I got another offer here as well, which would have paid only slightly less, and didn't involve Wesnoth.
20160804 21:47:17< crimson_penguin> But yeah, I mean, engineering jobs in NYC pay very well — maybe not quite as well as SF, but probably similar or more than Seattle
20160804 21:47:26-!- SigurdFD [~SigurdFD@dynamic-acs-72-23-110-58.zoominternet.net] has joined #wesnoth-dev
20160804 21:47:42< crimson_penguin> (And about double Vancouver)
20160804 21:48:49< vultraz> Interesting
20160804 21:49:02< vultraz> More than software designers?
20160804 21:58:18< crimson_penguin> Yeah, I think most than designers; you do mean actual design though, right? like, visual design?
20160804 22:00:41-!- ChipmunkV [~vova@d0017-2-88-172-31-68.fbx.proxad.net] has quit [Quit: ChipmunkV]
20160804 22:04:35-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 244 seconds]
20160804 22:17:04-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20160804 22:17:18-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160804 22:29:09-!- Appleman1234 [~Appleman1@KD036012024194.au-net.ne.jp] has joined #wesnoth-dev
20160804 22:31:49-!- atarocch [~atarocch@ipbcc2d6b3.dynamic.kabel-deutschland.de] has quit [Remote host closed the connection]
20160804 22:33:00-!- gfgtdf [~chatzilla@x4e369370.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 48.0/20160726073904]]
20160804 22:41:05< irker679> wesnoth: Celtic Minstrel wesnoth:wml_tag_porting f938b00246d0 / / (9 files in 4 dirs): Properly port [modify_side] to Lua https://github.com/wesnoth/wesnoth/commit/f938b00246d034a1d0109954218136ac684510c3
20160804 22:41:16< celmin> So, that's not quite finished yet, but any thoughts?
20160804 22:41:42< celmin> (Missing reset_map and I think one or two other things.)
20160804 22:51:38< celmin> (Also no testing has been done yet.)
20160804 22:55:55< vultraz> crimson_penguin: more like software engineers 
20160804 22:57:52-!- RatArmy [~RatArmy@om126161124111.8.openmobile.ne.jp] has quit [Ping timeout: 240 seconds]
20160804 23:00:03-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 240 seconds]
20160804 23:01:14< fabi_> vultraz: Are we still using wmllint to port addons between versions?
20160804 23:01:21< celmin> We are.
20160804 23:01:39< fabi_> And we still validate mainline with wmllint?
20160804 23:01:43< celmin> Well, rather, we're providing wmllint for addon devs to use in porting their addons.
20160804 23:01:45< celmin> I think so?
20160804 23:01:48< celmin> Why do you ask?
20160804 23:02:11< vultraz> We aren't
20160804 23:02:13< vultraz> Well
20160804 23:02:17< vultraz> At least no one has run it recently
20160804 23:03:06< fabi_> Then make a wmllint run of mainline part of what irker does after every commit.
20160804 23:03:20< celmin> Did you find something bad, or something?
20160804 23:03:50< fabi_> not yet
20160804 23:04:04< fabi_> Well, some small bugs
20160804 23:04:10< celmin> Well, it would be possible to make Travis run wmllint, I suppose.
20160804 23:04:14< fabi_> but nothing wmllint would catch
20160804 23:04:29< fabi_> yes, travis, not irker
20160804 23:05:03< vultraz> might want wmlindent too
20160804 23:05:08< vultraz> except not on the gui/ directory
20160804 23:05:17< fabi_> also wmlscope
20160804 23:05:35< celmin> vultraz: Note that if we make Travis run wmllint or whatever that it would be a "dry run", so wmlindent is probably a bit silly.\
20160804 23:06:02< fabi_> i think wmlindent has also a dry run option?
20160804 23:06:14< vultraz> wmlscope is useless
20160804 23:06:22< vultraz> burn it with fire.
20160804 23:06:26< Aginor> that's all sounding like good suggestions, let's add it to travis :)
20160804 23:06:28< celmin> Next you'll be asking to run clang-formatter… 
20160804 23:06:39< vultraz> if someone adds these things
20160804 23:06:44< celmin> (Or similar)
20160804 23:06:48< Aginor> celmin: good idea! :D
20160804 23:06:51< celmin> :|
20160804 23:06:54< vultraz> *please* update the gcc compiler to 4.8 while you're at it
20160804 23:07:46< Aginor> vultraz: got a strategy for security issues yet?
20160804 23:07:58< vultraz> Aginor: I have not.
20160804 23:08:13< vultraz> Besides "wait for GitHub to implement non-public issues"
20160804 23:08:21< fabi_> vultraz: Who took care about irker and travis before?
20160804 23:08:29< vultraz> iceiceice
20160804 23:08:34< vultraz> did travis
20160804 23:08:45< vultraz> irker... not sure.
20160804 23:08:47< celmin> I can probably do Travis stuff. I think Aginor and gfgtdf also have.
20160804 23:08:53< vultraz> irker was written in part by shadowm, IIRC.
20160804 23:08:58< vultraz> And e s r(?)
20160804 23:10:16< fabi_> vultraz: wmlscope worked at some time in the past. What is wrong with it nowadays?
20160804 23:10:23< vultraz> dunno
20160804 23:10:32< celmin> I thought it worked.
20160804 23:10:40< vultraz> I recall it couldn't handle something
20160804 23:10:43< celmin> Didn't someone run it on master not too long ago?
20160804 23:11:01< celmin> Maybe E_H fixed it when he upgraded all the Python tools to Py3?
20160804 23:14:30< fabi_> celmin: I asked because It would be nice if wmllint could be used to get rid of {FOREACH while porting.
20160804 23:14:51< celmin> Probably could. Ask E_H.
20160804 23:15:28< fabi_> In general, a tool which can expand hand picked macros would be nice.
20160804 23:15:41< fabi_> while preventing everything else in place, like comments
20160804 23:16:33< fabi_> Elvish_Hunter: ^
20160804 23:16:55-!- travis-ci [~travis-ci@ec2-54-91-37-166.compute-1.amazonaws.com] has joined #wesnoth-dev
20160804 23:16:56< travis-ci> wesnoth/wesnoth#10125 (wml_tag_porting - f938b00 : Celtic Minstrel): The build failed.
20160804 23:16:56< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/149896092
20160804 23:16:56-!- travis-ci [~travis-ci@ec2-54-91-37-166.compute-1.amazonaws.com] has left #wesnoth-dev []
20160804 23:18:06< celmin> Unit test failures, huh...
20160804 23:18:58< vultraz> blah.
20160804 23:19:03< celmin> Well, I see at least one that could be a result of forgetting to implement some part of [modify_side]
20160804 23:19:04< vultraz> this bug
20160804 23:19:19< celmin> BTW, you had no thoughts on that commit?
20160804 23:20:05< vultraz> why does color need a redraw?
20160804 23:20:25< celmin> Because it affects flags, I guess.
20160804 23:20:33< celmin> And TC for that matter.
20160804 23:20:48< celmin> So I guess all the flags and units for that side would need to be redrawn.
20160804 23:22:03< vultraz> I think it looks alright
20160804 23:22:22< fabi_> celmin: I wished everything actionWML would be wrapped in something like you did in [foreach] with [do].
20160804 23:22:34< celmin> fabi_: It pretty much is.
20160804 23:22:43< fabi_> sadly not in [event]
20160804 23:22:51< fabi_> were it is mixed with filter and other stuff
20160804 23:22:55< celmin> Well, I suppose you're right about [event], yeah.
20160804 23:23:06< celmin> Though you could wrap it in [command].
20160804 23:23:19< vultraz> celmin: you should add a deprecation warning about share_view/share_vision
20160804 23:23:32< vultraz> if this doesn't trigger the c++ one
20160804 23:23:52< celmin> Wait, was there a deprecation warning in the C++? I don't think I saw one.
20160804 23:24:02< fabi_> Yes, it can be wrapped in [command], sadly we never enforced that.
20160804 23:24:18< celmin> (Though actually, with this there shouldn't be one since you shouldn't get deprecations warnings from loading a saved game.)
20160804 23:24:59< vultraz> well there should be one somewhere
20160804 23:26:29-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 260 seconds]
20160804 23:27:52< celmin> Complain to gfgtdf then?
20160804 23:28:06< celmin> (But I can add one in the Lua implementation.)
20160804 23:31:16< celmin> I need to figure out how to handle the two shroud things.
20160804 23:31:32< irker679> wesnoth: Charles Dang wesnoth:master 1bb78a18f1b9 / src/gui/core/event/ (distributor.cpp handler.cpp): Expanded use of GUI2_SHOW_UNHANDLED_EVENT_WARNINGS guard https://github.com/wesnoth/wesnoth/commit/1bb78a18f1b95e2b73cfc54cab61a7d37101f2ed
20160804 23:31:44< vultraz> shut up, gui2
20160804 23:31:50< celmin> Haha
20160804 23:32:34< celmin> You know, you could've simply demoted it to debug level instead of adding guards. :P
20160804 23:35:52< vultraz> hmm
20160804 23:35:58< vultraz> under what circumstances would flag_rgb be empty..
20160804 23:36:16< celmin> Hmm?
20160804 23:36:50< vultraz> these ~(RC>1) warnings in the console sometimes
20160804 23:36:59< vultraz> happen when you select a save without a unit image to display
20160804 23:37:10< vultraz> I guess that would be when it's empty..
20160804 23:38:12< vultraz> but the unit class seems to imply it's sometimes empty
20160804 23:39:09< vultraz> not even sure why we have that key
20160804 23:39:14< vultraz> it's never anything BUT magenta 
20160804 23:39:42< vultraz> oh, huh, units has a TC_image_mods function
20160804 23:39:44< vultraz> I should make use of this
20160804 23:39:44< celmin> Is it ever magenta?
20160804 23:39:49< celmin> I'd expect it to be flag_green.
20160804 23:40:29< celmin> Unless I misremembered what it's for.
20160804 23:40:58< vultraz> huh
20160804 23:41:00< vultraz> hm
20160804 23:41:04< vultraz> what *is* it for..
20160804 23:41:19< celmin> Wiki
20160804 23:41:31< vultraz> unit::team_color() returns it..
20160804 23:41:40< celmin> Maybe it is magenta, I dunno
20160804 23:42:34< vultraz> "flag_rgb: usually set by MAGENTA_IS_THE_TEAM_COLOR; specifies the colours in the base flag to use for team-colouring the unit, expressed as a colour name (such as magenta) or a comma-separated list of RGB values (in hex format)."
20160804 23:42:50< celmin> Where is MAGENTA_IS_THE_TEAM_COLOR defined...
20160804 23:43:25< celmin> And if flag_rgb specifies the TC origin for the sprite, what specifies the TC origin for the flag or ellipse?
20160804 23:43:35< vultraz> macros/image-utils.cfg
20160804 23:43:47< celmin> Huh, it
20160804 23:43:51< celmin> 's that simple
20160804 23:44:55< vultraz> flag_rbg is set in the master game_config
20160804 23:45:01< vultraz> to flag_green
20160804 23:45:05< celmin> I see.
20160804 23:45:14< celmin> So if you think about it, flag_rgb is a misnomer.
20160804 23:45:23< vultraz> it is
20160804 23:45:30< celmin> I wonder if flag_rgb is ever actually used by UMC.
20160804 23:45:30< vultraz> honestly, I don't think there's any use for it
20160804 23:45:43< celmin> If it's not, we could change the name without any difficulty.
20160804 23:46:05< irker679> wesnoth: Charles Dang wesnoth:master 788671e45cdb / src/save_index.cpp: Don't try to set save index leader image TC if flag_rgb is empty https://github.com/wesnoth/wesnoth/commit/788671e45cdb08c26a64323bb02a8ce51e256821
20160804 23:46:17< vultraz> probably not the best fix
20160804 23:46:24< celmin> Eh, seems fine to me.
20160804 23:46:26< vultraz> might be better to have some check higher up
20160804 23:46:28< vultraz> but it works
20160804 23:51:00< vultraz> celmin: honestly, it *can't* really be any other value
20160804 23:51:06< vultraz> than magenta 
20160804 23:51:10< celmin> No, it can be flag_gree.
20160804 23:51:12< celmin> ^n
20160804 23:51:15< celmin> Or ellipse_red.
20160804 23:51:24< celmin> Of course, in practice, it never will be.
20160804 23:51:31< vultraz> a unit type with flag_green?
20160804 23:51:35< celmin> But it could be.
20160804 23:51:43< vultraz> but in practice not
20160804 23:51:50< vultraz> I think it could be safe to remove
20160804 23:51:54< vultraz> or move to game_config
20160804 23:52:04< celmin> Actually, thinking about it, it's totally useless to have flag_rgb as a key of [side].
20160804 23:52:25< celmin> Move to game_config then. Call it unit_rgb or something.
20160804 23:52:46< celmin> Maybe accept it in [unit_type] too.
20160804 23:52:46< vultraz> yes
20160804 23:52:50< vultraz> it already is
20160804 23:52:55< celmin> Ah, it is?
20160804 23:52:58< vultraz> that's where MAGENTA_IS_THE_TEAM_COLOR is used
20160804 23:53:03< celmin> Oh.
20160804 23:53:05< celmin> Duh.
20160804 23:53:06< celmin> Well.
20160804 23:53:08< vultraz> in every single *()@* unit type
20160804 23:53:19< vultraz> so this would be an improvement
20160804 23:53:31< celmin> Moving it to game_config is still fine - make that the default, but still accept it in [unit_type] for people actually wanting a different origin colour.
20160804 23:53:46< celmin> I wonder if the blue palette is still used anywhere...
20160804 23:53:56< vultraz> eh
20160804 23:54:02< vultraz> do we have to accept it in unit_type?
20160804 23:54:03< celmin> It wouldn't surprise me if the water doesn't conform to it at all.
20160804 23:54:18< celmin> Yes, I think we have to accept it in unit_type.
20160804 23:54:40< vultraz> I don't really see why
20160804 23:54:49< vultraz> All sprites are drawn with the magenta palette.
20160804 23:55:00< celmin> Can you prove this?
20160804 23:55:30< celmin> Including addons, of course.
20160804 23:55:40< vultraz> It's just... standard wesnoth convention.
20160804 23:55:59< celmin> Furthermore, why reduce the engine's flexibility for no reason?
20160804 23:56:00< vultraz> Without it, flag_rgb would have to be set to hex colors manually
20160804 23:57:24< celmin> Also, this isn't really a reason valid on its own, but… leaving it accepted in unit_type minimizes the code impact of the change.
20160804 23:57:34< celmin> You only need to update the unit_type class and the game_config.
20160804 23:57:51< vultraz> Uh...
20160804 23:58:00< celmin> Removing it from unit_type means you need to update every single place where it's used.
20160804 23:58:08< vultraz> I can't figure out why unit::team_color() returns flag_rgb_ :|
20160804 23:58:16< celmin> Isn't it obvious?
20160804 23:58:32< celmin> Hmm.
20160804 23:58:36< vultraz> not really?
20160804 23:58:40< vultraz> team colors are assigned to sides
20160804 23:58:42< vultraz> not units
20160804 23:58:51< celmin> Well, I suppose from the name team_color you might expect it to be the target colour, not the origin colour.
20160804 23:59:01< vultraz> exactly
20160804 23:59:04< celmin> Ah, but there you have it.
20160804 23:59:17< celmin> The unit's team colour is magenta, the side's team colour is red or whatever.
20160804 23:59:50< fabi_> there is also a color used as the magenta of the flags
20160804 23:59:57< fabi_> it is some green
20160804 23:59:58< celmin> flag_green, yes.
--- Log closed Fri Aug 05 00:00:03 2016