--- Log opened Fri Oct 28 00:00:09 2016
20161028 00:01:03-!- Shiki [~Shiki@141.39.226.226] has joined #wesnoth-dev
20161028 00:01:18-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 245 seconds]
20161028 00:04:24-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161028 00:13:02-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has quit [Disconnected by services]
20161028 00:13:33-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has joined #wesnoth-dev
20161028 00:13:36-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161028 00:23:44-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161028 00:27:06-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161028 00:32:20-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20161028 00:50:29-!- Shiki [~Shiki@141.39.226.226] has quit [Quit: Verlassend]
20161028 00:58:09-!- gfgtdf [~chatzilla@x4e36a218.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 49.0.2/20161019084923]]
20161028 01:03:48-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Read error: Connection reset by peer]
20161028 01:08:24-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20161028 01:48:11-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20161028 02:12:46< celticminstrel> vultraz: About team index refactor, I might have to just redo it now...
20161028 02:19:28-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20161028 02:19:34-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20161028 03:03:28< celticminstrel> There's a problem with the scroll wheel in the editor.
20161028 03:03:53< celticminstrel> When the mouse is over the palette, the wheel scrolls both the palette and the map.
20161028 03:04:05< Aginor> that's been there for a _very_ long time
20161028 03:04:14< vultraz> ^
20161028 03:05:03< aeth> the other problem I've noticed in the editor is that the giant box that tells you what terrains you're using (as if you can't tell by the icons already) block the view of the bottom border so you can't actually see what you're drawing until you either zoom way in/out or load the map in the game
20161028 03:05:03< celticminstrel> I never noticed it before...
20161028 03:05:10< aeth> both issues really bother me
20161028 03:05:19-!- JyrkiVesterinen [~JyrkiVest@87-100-131-6.bb.dnainternet.fi] has joined #wesnoth-dev
20161028 03:05:20< celticminstrel> What giant box?
20161028 03:05:28< celticminstrel> Oh.
20161028 03:05:32< celticminstrel> The tooltip-ish thing?
20161028 03:05:34< aeth> celticminstrel: hard to miss it if you're editing the bottom border
20161028 03:06:02< celticminstrel> Doesn't really block what you're drawing though...
20161028 03:06:17< aeth> it's "Extremely Long Terrain Name #1/Even Longer Name of the Secondary Terrain #2" in a giant font
20161028 03:06:27< aeth> it doesn't block your drawing but it blocks your view of it
20161028 03:06:37< celticminstrel> Not really?
20161028 03:07:05< aeth> on my screen, I can't see about half of the border terrains (in the bottom middle) at all
20161028 03:07:37< celticminstrel> But you can see the upper half, which is enough, isn't it?
20161028 03:07:39< aeth> the font in git is a bit smaller
20161028 03:07:53< aeth> celticminstrel: every other one you can see the upper half, every other one is completely obscured
20161028 03:07:53< celticminstrel> Oh, is it smaller on master?
20161028 03:08:03< aeth> it's narrower in master, but I think the same height
20161028 03:08:11< aeth> unless it's changed again since I last compiled master
20161028 03:08:19< celticminstrel> Oh wait, you're right.
20161028 03:09:21< aeth> And the worst part is that if you're used to Wesnoth you know exactly what terrains you're using from the icons in the center anyway
20161028 03:09:27< aeth> So it's completely useless for anyone who's not new to Wesnoth
20161028 03:10:09< celticminstrel> The settings button in the editor map generation dialog doesn't work properly.
20161028 03:10:34< celticminstrel> It shows the default generator settings even for a map not using the default generator.
20161028 03:10:43< celticminstrel> Maybe that'll be easy enough for me to fix...
20161028 03:10:47< vultraz> should be
20161028 03:11:17< celticminstrel> Hmm, clicking "generate" didn't actually use the selected generator, either...
20161028 03:11:27< vultraz> wha??
20161028 03:11:52< celticminstrel> It does if it's a default generator, but not if it's a Lua one.
20161028 03:11:57< vultraz> i see
20161028 03:12:27< celticminstrel> I guess for now I'll test in MP Create instead of the editor.
20161028 03:12:47< celticminstrel> I'll fix those issues later.
20161028 03:12:59< celticminstrel> Hmm...
20161028 03:13:04< vultraz> dont use 800x600!
20161028 03:13:12< celticminstrel> What has that got to do with anything?
20161028 03:14:17< celticminstrel> Wait, the reason it didn't work in the editor could be because there's an error in the generator?
20161028 03:14:36< celticminstrel> When I tried in MP Create I got "Mandatory WML child missing but untested for", whatever that means.
20161028 03:15:00< vultraz> ftr the cave generator doesn't appear in mp create anymore
20161028 03:15:01< celticminstrel> Then again... that might be more a GUI thing?
20161028 03:15:06< celticminstrel> I know.
20161028 03:15:12< celticminstrel> (Wait, it used to?
20161028 03:15:13< celticminstrel> )
20161028 03:15:16< vultraz> yes
20161028 03:15:41< celticminstrel> So why are you telling me not to use 800x600?
20161028 03:15:51< vultraz> because you always use 800x600
20161028 03:15:56< celticminstrel> That's not a reason.
20161028 03:17:15< celticminstrel> Well, it doesn't seem to work in MP Create, either...
20161028 03:17:30-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161028 03:18:20< celticminstrel> It doesn't seem like the Lua generator is ever actually called...
20161028 03:18:44< celticminstrel> Do none of SFD's random maps use Lua?
20161028 03:19:19 * celticminstrel wouldn't be all that surprised since the default generator is pretty flexible.
20161028 03:19:26< vultraz> dont think so
20161028 03:23:00< celticminstrel> Hmm, it doesn't even work in the GUI1 MP Create... fails with the same error, which makes me wonder if there's a problem with my WML...
20161028 03:24:04< vultraz> certain keys are expected
20161028 03:24:11< vultraz> look in the create engine
20161028 03:24:13-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161028 03:24:31< celticminstrel> Oh, maybe name is required?
20161028 03:25:08< vultraz> probably
20161028 03:26:28< celticminstrel> This error message is so useless though...
20161028 03:26:37< celticminstrel> I need the debugger to properly track it. :/
20161028 03:27:10< JyrkiVesterinen> IIRC, the message is thrown by code that simply doesn't have much more information available.
20161028 03:27:20< celticminstrel> Yeah.
20161028 03:27:43< celticminstrel> There's no way to make the error message more useful at the point where it's thrown.
20161028 03:27:56< celticminstrel> Because it's thrown from the config child that doesn't exist, in effect.
20161028 03:28:14< celticminstrel> Maybe it would help to get rid of the null config and bool operator, though.
20161028 03:28:29< celticminstrel> I think that's a major change though.
20161028 03:28:42-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 256 seconds]
20161028 03:29:23< JyrkiVesterinen> I think it would be better design if config::child() threw an exception if the child doesn't exist. Calling code would then catch the exception and print an appropriate error message.
20161028 03:29:34< celticminstrel> That sounds like a good idea.
20161028 03:29:50< JyrkiVesterinen> But it would indeed be a major change. Forget a call site and the game suddenly crashes when it encounters invalid WML.
20161028 03:29:58< celticminstrel> Though if we did that it wouldn't automatically force the calling code to update.
20161028 03:30:02< celticminstrel> Yeah, what you said.
20161028 03:33:06< celticminstrel> ...I'm not getting the error anymore. :|
20161028 03:34:20< celticminstrel> Maybe that means it was name after all? Though I used F5... so I dunno why restarting Wesnoth made a difference...
20161028 03:35:57< vultraz> create engine not recreated or something?
20161028 03:36:12< celticminstrel> I should make the mapgen kernel apply a metatable to the params table that makes it give a better error message.
20161028 03:36:22< celticminstrel> Oh wait... that wouldn't work for optional params... hmm...
20161028 03:36:44< celticminstrel> Would village density be a required parameter or should it have a default?
20161028 03:37:47< celticminstrel> vultraz: Isn't the create engine created when the dialog is opened? If you go to the title screen it's destroyed, right?
20161028 03:38:22< celticminstrel> It seems that when you select a random map, any map generator errors are suppressed.
20161028 03:38:35< celticminstrel> You don't see them unless you actually try to advance to the next screen.
20161028 03:38:54< celticminstrel> Which is actually kind of weird, since it seems to imply that the map you see won't be the one used for the actual game...
20161028 03:39:46< celticminstrel> Well, that doesn't seem to be the case though, at least with the default generator.
20161028 03:43:03< celticminstrel> Oh, I think I figured out why the map display is blank. I used an invalid terrain for the wall.
20161028 03:46:33< celticminstrel> Hmm, looks like it sort of works now, but... that cave can't be of size 20...
20161028 03:54:43< celticminstrel> Oh, I see a problem. helper.rand specifies ranges as .. but the old cave generator specified them as -
20161028 03:55:08< celticminstrel> I guess it's fine to change it though. It's not like compatibility with the old cave generator is needed.
20161028 03:59:47< celticminstrel> Ah... it's a caught exception so I need a bit more work to find it in the debugger...
20161028 04:10:45< celticminstrel> So it's thrown by the dependency manager, in get_required...
20161028 04:16:41< celticminstrel> It seems to think the scenario has no ID, but it does...
20161028 04:26:57-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Quit: wedge009]
20161028 04:27:14-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161028 04:46:24-!- Kwandulin [~Miranda@p5DDD2B8F.dip0.t-ipconnect.de] has joined #wesnoth-dev
20161028 04:58:44-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20161028 05:12:28-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161028 05:17:04-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20161028 05:59:56-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20161028 06:05:05-!- JyrkiVesterinen [~JyrkiVest@87-100-131-6.bb.dnainternet.fi] has quit [Quit: .]
20161028 06:17:51-!- atarocch [~atarocch@93.56.160.28] has joined #wesnoth-dev
20161028 06:22:09-!- atarocch_ [~atarocch@93.56.160.28] has joined #wesnoth-dev
20161028 06:22:40-!- atarocch [~atarocch@93.56.160.28] has quit [Ping timeout: 265 seconds]
20161028 06:42:50-!- VultCave [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20161028 06:42:50-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection reset by peer]
20161028 06:43:07-!- VultCave is now known as vultraz
20161028 07:00:53-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161028 07:05:03-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 245 seconds]
20161028 07:11:10-!- JyrkiVesterinen [~JyrkiVest@194.157.54.14] has joined #wesnoth-dev
20161028 07:32:22-!- JyrkiVesterinen [~JyrkiVest@194.157.54.14] has quit [Quit: .]
20161028 08:00:47-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
20161028 08:13:21-!- Nobun [~nobun@5.170.108.236] has joined #wesnoth-dev
20161028 08:15:07-!- Kwandulin [~Miranda@p5DDD2B8F.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20161028 08:17:08-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Quit: wedge009]
20161028 08:17:29-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161028 08:21:20-!- Nobun [~nobun@5.170.108.236] has quit [Ping timeout: 260 seconds]
20161028 08:48:53-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Quit: wedge009]
20161028 08:49:05-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161028 08:49:13-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161028 08:51:57-!- Kwandulin [~Miranda@p5DDD2B8F.dip0.t-ipconnect.de] has joined #wesnoth-dev
20161028 08:53:26-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 250 seconds]
20161028 09:14:44-!- Ivanovic [~ivanovic@p579FBF3F.dip0.t-ipconnect.de] has quit [Changing host]
20161028 09:14:44-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20161028 09:47:47< vultraz> DeFender1031: did you get your clone cloned?
20161028 09:50:37< DeFender1031> vultraz, haven't tried yet, i'm stuck on the build requisites because i've got a weird debian setup and i'm stuck with a bunch of conflicting packages and can't get sdl2 to install :(
20161028 09:50:52< vultraz> oh dear :(
20161028 09:52:31< DeFender1031> truth is, for this one function i'm working on, i could probably just rip out the code and test it and rewrite it independently, but still, i'f i'm going to be serious about this, i'm not going to be able to do that with everything
20161028 09:55:50< vultraz> yeah
20161028 09:56:04< vultraz> and we always need new contributors :)
20161028 09:57:58< DeFender1031> part of the problem is that KDE4 killed off a bunch of the programs that i use regularly, and therefore the packages haven't been maintained and if I try to upgrade many things, it tries to remove them. I really need to get around to trying out trinity (an independent fork of KDE3 that IS maintained and supposedly has all the features of it) except that the only way to upgrade would be killing off what I have, and then if it 
20161028 09:58:00< DeFender1031> doesn't work, there's no way to revert and I'm screwed. I might actually just set up a new partition, install it there, make sure it's all got what i need, and then upgrade... It's a hassle, but this isn't the first time i've found myself stuck because of this. I usually just give up and say "whatever, I don't actually care enough" but I really REALLY want to work on this, so I think it's time. Sigh...
20161028 09:58:57< vultraz> well, shadowm is also a debian user, so ping him if you need help. I can't offer any assistance, I'm afraid :P
20161028 10:00:17< vultraz> it's been ages since I've used linux, and even then it was ubuntu 
20161028 10:00:31< DeFender1031> it's unlikely anyone but me and one friend of mine have this particular issue.
20161028 10:00:48< vultraz> (I wonder if ubuntu's UI is any better from 12.10)
20161028 10:00:55< vultraz> current ui*
20161028 10:00:57< DeFender1031> (But I will most certainly be pestering that one friend.)
20161028 10:02:10< vultraz> you don't have a windows vm or something lying around?
20161028 10:16:29-!- Kwandulin [~Miranda@p5DDD2B8F.dip0.t-ipconnect.de] has quit [Ping timeout: 260 seconds]
20161028 10:37:23-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161028 10:41:49-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 252 seconds]
20161028 10:46:23< DeFender1031> vultraz, I could do it in VM, but developing in VM is very painful.
20161028 10:46:50< DeFender1031> I'd rather just get off my ass and fix my main machine.
20161028 10:46:59< vultraz> heh
20161028 10:53:14< DeFender1031> well, in the meantime, i'm going to continue to get my local repo set up. I'm currently cloning from master, after which i'll configure my push URL to be my own fork
20161028 10:56:07< vultraz> one usually sets up 2 remotes
20161028 10:56:09< vultraz> origin and form
20161028 10:56:11< vultraz> fork*
20161028 10:56:18< vultraz> then pushes to each as appropriate 
20161028 10:56:21< vultraz> origin is automatically there
20161028 10:56:26< DeFender1031> right
20161028 10:57:32< DeFender1031> but in my case, i don't expect anyone to push to my fork but me, and therefore there will be nothing to pull from it, and I don't have push access for main. If that changes, I can always set up an additional push url then
20161028 11:00:06< DeFender1031> I mean, ideally, if github had a system to submit PRs from a local clone, i'd not even HAVE a github fork and just develop locally and submit my code directly for review. It seems like the only reason I need my own fork is because PRs are pull requests, meaning requests to pull changes from this repo into that, rather than push requests, meaning requests to accept code that this repo has SENT that one.
20161028 11:00:19< DeFender1031> unless i'm still totally misunderstanding the system
20161028 11:00:24< DeFender1031> which is always possible.
20161028 11:08:16< vultraz> pull requests are essentially 'will the main repo please accept these commits that are on the fork branch but not in the origin branch'
20161028 11:09:01< vultraz> so say you push to 'my bugfix' on your fork, you can open a PR against origin/master (via the web interface, ofc), that will ask us to accept any commits on your 'by bugfix' branch not in master
20161028 11:09:19< DeFender1031> right. if they were instead "here, main repo, have some code, put it into a probationary area pending acceptance" then i wouldn't even need my own fork.
20161028 11:09:33< vultraz> yeah, that's not how it works
20161028 11:09:47< DeFender1031> right. that's what i said.
20161028 11:09:48< vultraz> unless you have push access and push to a new branch of the main repo :)
20161028 11:10:01< DeFender1031> right, but then it's not REALLY probationary
20161028 11:10:09< vultraz> pull requests are essentially a request to merge a branch
20161028 11:10:14< DeFender1031> right
20161028 11:10:17< vultraz> branches are a core got workflow concept
20161028 11:10:20< vultraz> git*
20161028 11:10:37< DeFender1031> and since the main repo can pull from my local machine, i need a github fork for it to pull from
20161028 11:10:40< vultraz> so there's no magic "staging area"
20161028 11:11:12< DeFender1031> well... there IS a "staging area", but that's for commits and is different.
20161028 11:11:26< vultraz> (technically such a thing exists in every local clone called FETCH_HEAD, which is a special internal branch that stores changes if you use git fetch instead of git pull)
20161028 11:11:39< DeFender1031> huh....
20161028 11:12:43< vultraz> git pull just attempts to get changes and automatically creates a merge commit
20161028 11:12:51< vultraz> a merge commit is a commit where two branches are merged
20161028 11:12:58< vultraz> we usually try to avoid merge commits
20161028 11:13:03< vultraz> and rebase them out
20161028 11:13:18< vultraz> using git fetch followed by git merge is an implicit rebase
20161028 11:13:18< DeFender1031> rebase?
20161028 11:13:20-!- JyrkiVesterinen [~JyrkiVest@194.157.54.14] has joined #wesnoth-dev
20161028 11:13:27< DeFender1031> ah
20161028 11:13:33< vultraz> rebasing is when two branches are made to match and the difference applied on top
20161028 11:13:36< DeFender1031> yeah, i more or less understand all of this
20161028 11:13:58< vultraz> so say you have branch A and branch B. branch B is 1 commit ahead, 5 commits behind master
20161028 11:14:02< vultraz> (branch A)
20161028 11:14:31< DeFender1031> all i was saying is that the only purpose the fork really serves is to act as a staging area for PRs for main.
20161028 11:14:47< vultraz> if you rebase branch B onto branch A, what happens is those 5 commits are applied to branch B and a *new* commit derived from that '1 commit ahead' is applied on top
20161028 11:14:51< vultraz> without a merge commit
20161028 11:14:54< DeFender1031> no one is going to be making PRs into Deffy's random fork.
20161028 11:15:10< vultraz> Well, in this case.
20161028 11:15:18< DeFender1031> right, that's what i mean.
20161028 11:15:43< DeFender1031> in general, obviously anyone can make PRs into any fork and any fork can include or not include branche from any other fork
20161028 11:15:52< vultraz> Only you have push access to your fork anyway, since it's on your account :P
20161028 11:16:02< DeFender1031> exactly
20161028 11:17:32< vultraz> An important thing to remember is that *any commits* to a branch in an open PR are counted as part of that PR
20161028 11:17:56< vultraz> so if you want to open two simultaneous PRs, you need to use 2 different branches *based on their common ancestor*
20161028 11:18:28< vultraz> say you have 2 commits in 2 prs. You'll need to make branch_foo and branch_bar, both created from master.
20161028 11:18:43< vultraz> if you create branch_foo, add a commit, open a PR, and then create branch_bar from there
20161028 11:18:56< vultraz> your second PR will include both commits
20161028 11:19:05< vultraz> (assuming you're PRing against master, that is)
20161028 11:21:14< vultraz> So just remember, any commits to a branch that you're requesting to merge via PR count as part of the PR
20161028 11:21:22< vultraz> (and you can add more after the PR is open, don't worry)
20161028 11:22:25-!- Kwandulin [~Miranda@p5DDD2B8F.dip0.t-ipconnect.de] has joined #wesnoth-dev
20161028 11:23:47< vultraz> One other thing to keep in mind is that we usually ask you to *squash* your commits if you have a PR with a lot of merge commits or 'wip 1', 'wip2', etc
20161028 11:23:59< vultraz> squashing happens during rebasing 
20161028 11:24:09< vultraz> and it' when you squash multiple commits into one
20161028 11:24:41< vultraz> (during rebasing you can also *skip* commits, meaning they're ignored as part of the rebase. One should always skip merge commits when rebasing)
20161028 11:24:52< vultraz> it's*
20161028 11:25:24< vultraz> (it's also possible to *split* commits too)
20161028 11:27:16< vultraz> And one very important git-ism to remember is that commits are squashed with the one *below* it
20161028 11:28:15< vultraz> When rebasing, one can set a commit's status to either PICK, SKIP, SPLIT, or SQUASH
20161028 11:28:29< vultraz> You can also *reorder the commits*
20161028 11:28:36< vultraz> So say you have 4 commits: A B C and D
20161028 11:29:01< vultraz> you want to squash B into A. To do so, you set A to PICK (default) and B to SQUASH
20161028 11:29:06< vultraz> Easy
20161028 11:29:13< vultraz> But say you want to squash D into B
20161028 11:29:48< DeFender1031> keep talking, i'm still here, i'm just on the phone with my day job currently
20161028 11:29:50< vultraz> well, then you need to move (in whatever method whatever client you're using provides) D right above B, so the order reads A B D C
20161028 11:30:06< vultraz> and mark D as SQUASH, and B as PICK
20161028 11:30:29< vultraz> if one were to mark D as SQUASH *without* moving it, you'd end up squashing with C.
20161028 11:31:06< vultraz> But what if you want to squash everything into A?
20161028 11:31:17< vultraz> Then one just marks everything as SQUASH except A
20161028 11:31:21< vultraz> and everything goes into A :)
20161028 11:32:34< vultraz> the reason the base must be PICK is that adjacent SQUASHes are merged together
20161028 11:40:57< DeFender1031> [10/28/2016 2:51:36 pm] <vultraz> So just remember, any commits to a branch that you're requesting to merge via PR count as part of the PR
20161028 11:40:58< DeFender1031> [10/28/2016 2:51:44 pm] <vultraz> (and you can add more after the PR is open, don't worry)
20161028 11:41:20< DeFender1031> do you mean that if i add commits after opening the PR, that those commits will or will not be included in he PR if accepted?
20161028 11:42:01< vultraz> If you push more commits to the same branch they show up in the PR, yes.
20161028 11:42:56< DeFender1031> and if the PR is accepted before i made the additional commits?
20161028 11:43:09< vultraz> you'll need to open a new PR.
20161028 11:43:26< DeFender1031> heh... clone is at 12% after 50 minutes.
20161028 11:43:33< DeFender1031> okay
20161028 11:54:24-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 250 seconds]
20161028 12:10:16-!- tomreyn_ [~tomreyn@megaglest/team/tomreyn] has joined #wesnoth-dev
20161028 12:15:16-!- ShikadiLord [~ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20161028 12:15:24-!- ShikadiL1rd [~ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20161028 12:16:21-!- Kwandulin [~Miranda@p5DDD2B8F.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20161028 12:20:16-!- abruanese_ [~a@45.63.76.107] has joined #wesnoth-dev
20161028 12:20:50-!- Netsplit *.net <-> *.split quits: abruanese, shadowm, crimson_penguin, DeFender1031, DDR, prkc, timotei, tomreyn, shadowm_vps
20161028 12:25:43-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161028 12:32:52-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20161028 12:36:05-!- Netsplit over, joins: timotei, prkc
20161028 12:36:35-!- Netsplit over, joins: DDR
20161028 12:36:41-!- Netsplit over, joins: DeFender1031
20161028 12:37:29-!- crimson_penguin [~crimson_p@ec2.happyspork.com] has joined #wesnoth-dev
20161028 12:51:32< DeFender1031> vultraz, what are the commands for this squashing?
20161028 12:52:01< DeFender1031> (I asked a while ago but apparently that message didn't make it in before the netsplit)
20161028 12:52:42-!- prkc [~prkc@gateway/vpn/privateinternetaccess/prkc] has quit [Remote host closed the connection]
20161028 12:54:08< vultraz> it's done during git rebase
20161028 12:54:18< vultraz> but i don't know the commands since I only use a UI
20161028 12:54:52< JyrkiVesterinen> git rebase -i HEAD~4
20161028 12:55:14< JyrkiVesterinen> (if you want to squash/edit/reorder the last four commits)
20161028 12:55:56< JyrkiVesterinen> Squashing is part of history rewriting. Here is official reference about history rewriting: https://git-scm.com/book/en/v2/Git-Tools-Rewriting-History
20161028 12:56:34< JyrkiVesterinen> Also, history rewriting is considered an advanced topic. You can skip squashing if you feel it's too hard.
20161028 13:07:03-!- vultraz changed the topic of #wesnoth-dev to: 1.13.6 early November maybe? | Wesnoth Developers Channel | >>> Want to help? Go here: http://r.wesnoth.org/t42911 (and thanks!) <<< | Logs: http://irclogs.wesnoth.org | Bug tracker: http://bugs.wesnoth.org
20161028 13:13:23-!- louis94 [~~louis94@91.178.241.155] has joined #wesnoth-dev
20161028 13:19:55< DeFender1031> JyrkiVesterinen, good to know, thanks. I might skip it at first.
20161028 13:31:08-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161028 13:39:38-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 245 seconds]
20161028 13:39:47-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161028 13:40:42< DeFender1031> clone is at 30% after 2 hours and 48 minutes...
20161028 13:41:03< DeFender1031> there's no way to resume a clone if i stop it in the middle is there?
20161028 13:42:57-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161028 13:44:28< JyrkiVesterinen> DeFender1031: indeed, pausing a clone and resuming it later isn't possible.
20161028 13:44:29< JyrkiVesterinen> http://stackoverflow.com/questions/3954852/how-to-complete-a-git-clone-for-a-big-project-on-an-unstable-connection
20161028 13:45:18< JyrkiVesterinen> (You seem to have a quite slow connection. I don't recall exactly how long cloning the repo took for me, but I think it was maybe an hour or two.)
20161028 13:45:56< DeFender1031> 40 MBPS is the highest i can get where I am...
20161028 13:46:27< JyrkiVesterinen> I have 10 Mbps.
20161028 13:46:45< DeFender1031> odd
20161028 13:47:07< DeFender1031> then maybe the routing to GH is just slower... i don't know
20161028 13:47:55< DeFender1031> hmm... yeah. It's telling me that it's cloning at roughly 170 KiB/s
20161028 13:59:50-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 252 seconds]
20161028 14:04:49-!- JyrkiVesterinen [~JyrkiVest@194.157.54.14] has quit [Quit: Going home, and going to be offline for about three hours.]
20161028 14:08:08-!- louis94 [~~louis94@91.178.241.155] has quit [Ping timeout: 260 seconds]
20161028 14:30:42-!- crimson_penguin [~crimson_p@ec2.happyspork.com] has quit [Changing host]
20161028 14:30:42-!- crimson_penguin [~crimson_p@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20161028 14:32:41-!- DeFender1031 [~DeFender1@46-116-17-86.bb.netvision.net.il] has quit [Quit: I'm not back now.]
20161028 14:54:25-!- Kwandulin [~Miranda@p5DDD2B8F.dip0.t-ipconnect.de] has joined #wesnoth-dev
20161028 15:07:17-!- louis94 [~~louis94@91.178.241.155] has joined #wesnoth-dev
20161028 15:36:19-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161028 15:39:57-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 265 seconds]
20161028 15:52:49-!- gfgtdf [~chatzilla@x4e3638a9.dyn.telefonica.de] has joined #wesnoth-dev
20161028 15:52:56< gfgtdf> 20161028 03:29:23< JyrkiVesterinen> I think it would be better design if config::child() threw an exception if the child doesn't exist. Calling code would then catch the exception and print an appropriate error message.
20161028 15:53:39< gfgtdf> JyrkiVesterinen: this would break  all code that use if(config& cfg = c.child("tagname")) { /*act on tagname*/}
20161028 15:56:47< gfgtdf> celmin: there are already bugreports abotu the editor scrolling issue: http://gna.org/bugs/index.php?24501
20161028 16:06:06< gfgtdf> 20161028 03:15:00< vultraz> ftr the cave generator doesn't appear in mp create anymore
20161028 16:07:33< gfgtdf> vultraz: i rmvoed the test cavegen mp scneario when i made Htt use te lua cavegen since that mp scenario was just added for testing the lua cavegen in the first place.
20161028 16:07:49< vultraz> i see
20161028 16:08:08< gfgtdf> vultraz: also note that it was always in '#ifdef DEBUG_MODE'
20161028 16:08:22-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161028 16:09:38-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161028 16:09:44-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161028 16:09:56-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has quit [Ping timeout: 260 seconds]
20161028 16:15:36-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20161028 16:19:44-!- louis94 [~~louis94@91.178.241.155] has quit [Ping timeout: 260 seconds]
20161028 16:35:47-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20161028 16:35:53-!- abruanese_ [~a@45.63.76.107] has quit [Ping timeout: 260 seconds]
20161028 16:39:41-!- Jetrel [~Jetrel@2001:558:6014:1e:2422:435:dd84:bbf3] has quit [Read error: Network is unreachable]
20161028 16:40:29-!- Jetrel [~Jetrel@2001:558:6014:1e:2422:435:dd84:bbf3] has joined #wesnoth-dev
20161028 16:44:08-!- JyrkiVesterinen [~JyrkiVest@87-100-133-159.bb.dnainternet.fi] has joined #wesnoth-dev
20161028 16:46:57< JyrkiVesterinen> 20161028 15:53:39< gfgtdf> JyrkiVesterinen: this would break  all code that use if(config& cfg = c.child("tagname")) { /*act on tagname*/}
20161028 16:47:26< JyrkiVesterinen> For that, I could imagine something like "config* config::optional_child(std::string name)".
20161028 16:47:50< JyrkiVesterinen> My remark was a hypothetical one and not an actual suggestion, though.
20161028 16:48:06< vultraz> there's also if(!child_or_empty(name).empty()) {}
20161028 16:52:35< celticminstrel> The code using if(config& cfg = c.child("tagname")) is bad anyway.
20161028 16:53:08< gfgtdf> celticminstrel: why woudl thatbbe bad ?
20161028 16:53:26< celticminstrel> Because the bool operator in the config class is bad.
20161028 16:54:51< gfgtdf> celticminstrel: i think the bool operator is fine
20161028 16:55:18< celticminstrel> It compares the address of the config to a specific "magic" instance.
20161028 16:55:32< celticminstrel> That's not really good practice.
20161028 16:55:48< JyrkiVesterinen> The bool operator exists because config::child() doesn't throw exceptions. It has no way to report that the child doesn't exist other than returning the null config.
20161028 16:56:09< JyrkiVesterinen> It isn't good design. I think it's too late to change the design, mind you. :(
20161028 16:56:19< gfgtdf> celticminstrel: the boo operatoers was desogned to make "if(config& cfg = c.child("tagname")) " work.
20161028 16:56:23< celticminstrel> If it returned a pointer, that would be better, though still not ideal.
20161028 16:56:23< gfgtdf> bool*
20161028 16:56:31< vultraz> it always confuses me that config references can be used in if blocks like pointers
20161028 16:56:58< celticminstrel> gfgtdf: Really, what you should be doing is if(c.has_child("tagname")) {do stuff with c.child("tagname")}
20161028 16:57:31< gfgtdf> celticminstrel: thats ugly code, plus you have to lookup "tagname" twiche with that.
20161028 16:57:50< JyrkiVesterinen> I think celticminstrel's code is less ugly.
20161028 16:57:54< celticminstrel> gfgtdf: It's a map, so that shouldn't really be a problem.
20161028 16:57:54-!- Kwandulin [~Miranda@p5DDD2B8F.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20161028 16:58:04< JyrkiVesterinen> The existing idiom *looks* wrong.
20161028 16:58:16< vultraz> yes
20161028 16:58:17< vultraz> it does
20161028 16:58:19< celticminstrel> It looks wrong mainly because of the &
20161028 16:58:27< celticminstrel> If it was a pointer it wouldn't look so wrong.
20161028 16:58:28< vultraz> I try to use has_child
20161028 16:58:36< vultraz> what celticminstrel said
20161028 16:58:58< gfgtdf> it is a prpobem fpr multiple reasons reasons: 1) its slower, 2) you dublich cate the code by specifyint the tagname twice.
20161028 16:59:26< celticminstrel> 1) Technically it's slower sure, but I doubt that's significant in the majority of cases.
20161028 16:59:49< celticminstrel> It's O(log n) lookup, right?
20161028 17:00:14< celticminstrel> And in this case, n is the number of distinct child tags in the config. Multiple tags with the same name only count once.
20161028 17:00:17-!- ShikadiL1rd is now known as shadowm
20161028 17:00:45< celticminstrel> So if you have 100 child [tagname] and no other tags, it's constant time lookup (only one entry in the map).
20161028 17:00:54< gfgtdf> celticminstrel: yes but evne if its just one chodl tad, its requires copyign strings compaing strings, 
20161028 17:01:09< gfgtdf> celticminstrel: constats-time doesnt actualy imply its fast ina situation liek this.
20161028 17:01:13< gfgtdf> constant*
20161028 17:01:26< celticminstrel> Copying and comparing short strings.
20161028 17:01:34-!- stikonas_ is now known as stikonas
20161028 17:01:56< celticminstrel> We're not working with a novel here. It's not going to be significantly slower with an extra copy.
20161028 17:03:31< Soliton> unless we'd be talking about a class used everywhere in wesnoth.
20161028 17:05:07-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has joined #wesnoth-dev
20161028 17:09:00< celticminstrel> I dunno how much that matters...
20161028 17:09:51< celticminstrel> Anyway, if you're really worried about speed you can use child_or_empty.
20161028 17:10:02< JyrkiVesterinen> The "if (config& cfg = c.child("tagname")) {" idiom can be used in actual performance-critical code.
20161028 17:10:16< JyrkiVesterinen> But has_child() check should be used by default IMO.
20161028 17:10:39< celticminstrel> Even then I'd prefer if it was a pointer.
20161028 17:11:26< Soliton> you'd be reversing what the one introducing the current design did then.
20161028 17:11:44< celticminstrel> Hm?
20161028 17:11:45< JyrkiVesterinen> Regarding performance concerns, I'll just link here: https://wiki.wesnoth.org/HackingWesnoth#Guideline_1:_Don.27t_prematurely_optimize
20161028 17:12:37< Soliton> don't prematurely pessimize based on "looks wrong".
20161028 17:13:22< JyrkiVesterinen> I disagree with that.
20161028 17:13:54< JyrkiVesterinen> Code readability and ease to understand it should be the #1 priority unless there really is a need to optimize.
20161028 17:14:58< Soliton> i agree.
20161028 17:15:06< gfgtdf> JyrkiVesterinen: that code is easiyl understandable, the delaration in if staements if a ell knowon c++ conectpt.
20161028 17:15:21< gfgtdf> is a well*
20161028 17:15:33< JyrkiVesterinen> It is a well known C++ concept... with pointers, not references.
20161028 17:15:48< JyrkiVesterinen> With references it looks like a newbie C++ programmer mistake.
20161028 17:16:45-!- vultraz [~chatzilla@124.109.10.167] has quit [Remote host closed the connection]
20161028 17:16:48< Soliton> somehow there seems to be the assumption that the current design was implemented by a newbie c++ programmer on a whim.
20161028 17:17:34< JyrkiVesterinen> I do 
20161028 17:17:40< JyrkiVesterinen> not assume that way.
20161028 17:18:13< JyrkiVesterinen> (Asterisk and Enter are next to each other. I hit Enter when I was trying to write *not* with asterisks.)
20161028 17:19:06 * celticminstrel uses underline for that sort of thing.
20161028 17:19:31< JyrkiVesterinen> I can understand what the programmer was trying to do. Make the get_child accessor return the child, and signal the "no such child" error condition in some way.
20161028 17:20:12< JyrkiVesterinen> It's just not what I would do. I'd use an exception for that, and a separate method that returns a pointer (and nullptr if the child doesn't exist).
20161028 17:21:07< JyrkiVesterinen> I support the "fail-fast" way of error handling. If something is wrong, crash the program as quickly as possible, instead of attempting to continue and risking the potential to malfunction.
20161028 17:21:55< JyrkiVesterinen> I think the author of config::child() supports the "defensive programming" way instead, where the code tries to keep the program from crashing even if something is wrong.
20161028 17:22:24< JyrkiVesterinen> Neither exceptions, nor null pointers, are really compatible with defensive programming.
20161028 17:22:43< JyrkiVesterinen> Failing to catch an exception or to check a pointer for null will crash the program.
20161028 17:23:30< JyrkiVesterinen> (Dereferencing a null pointer is technically undefined behavior, but in practice it always causes a crash, save for some obscure corner cases.)
20161028 17:23:50< JyrkiVesterinen> And thus we got the null config and the bool operator.
20161028 17:24:24-!- Kwandulin [~Miranda@p5DDD2B8F.dip0.t-ipconnect.de] has joined #wesnoth-dev
20161028 17:24:48< celticminstrel> We have exception handlers in main and stuff though, so many exceptions are automatically caught eventually and converted to an error dialog. And if child() threw an exception, the error message could be made more useful than it is now.
20161028 17:25:08< celticminstrel> Specifically, it could give the name of the missing child.
20161028 17:25:30< celticminstrel> find_child() could also throw an exception with the same info plus the required attribute and value.
20161028 17:25:45-!- louis94 [~~louis94@91.178.241.155] has joined #wesnoth-dev
20161028 17:26:14< JyrkiVesterinen> That's the power of exceptions. Error conditions can be handled in the appropriate place instead of wherever the error condition is detected.
20161028 17:26:25< JyrkiVesterinen> Using the null config makes that impossible.
20161028 17:32:43-!- mjs-de [~mjs-de@x4e30c8e0.dyn.telefonica.de] has joined #wesnoth-dev
20161028 17:33:31< Soliton> should be fairly simple to test those theories and apply that generic advice.
20161028 17:34:05-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20161028 17:34:38< JyrkiVesterinen> What do you mean "fairly simple"? If we did change config::child() to throw exceptions, we would be almost guaranteed to miss places where the exception needs to be catched.
20161028 17:34:45< JyrkiVesterinen> It's too late to change now. :(
20161028 17:35:07< Soliton> fail fast is good, remember?
20161028 17:35:16-!- louis94 [~~louis94@91.178.241.155] has quit [Ping timeout: 260 seconds]
20161028 17:35:49< JyrkiVesterinen> It's how code should be initially written IMO.
20161028 17:35:54< celticminstrel> We would probably miss places, but then it would probably be caught anyway.
20161028 17:36:07-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20161028 17:36:32< celticminstrel> Changing it to return a pointer is actually easier, because that forces you to update every call site.
20161028 17:36:49< celticminstrel> Though I do prefer an exception over a pointer...
20161028 17:36:56< JyrkiVesterinen> But converting an existing codebase from defensive programming to fail-fast is almost impossible due to the sheer amount of code that relies on all these null pointer checks and whatever.
20161028 17:37:04< JyrkiVesterinen> (The reverse is easy, by the way.)
20161028 17:37:58< Soliton> the defensive programming was a guess from you, btw. i highly doubt it.
20161028 17:38:20-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20161028 17:38:32< JyrkiVesterinen> Well, what is your explanation for the existence of the null config in that case?
20161028 17:39:16< Soliton> gfgtdf explained that, no?
20161028 17:39:36< Soliton> to support the newbie c++.
20161028 17:40:15< JyrkiVesterinen> That you can do "if (config& cfg = c.child("tagname")) {" instead of "if (config* cfg = c.child("tagname")) {"?
20161028 17:41:19< JyrkiVesterinen> It's quite strange that someone would do all that effort in order to allow programmers to get the child by reference instead of by pointer.
20161028 17:41:43< JyrkiVesterinen> Especially when there is the cost of making the resulting code extremely non-idiomatic.
20161028 17:43:11< vultraz> what's wrong with simply dereferencing the pointer ;|
20161028 17:43:52< vultraz> or just using ->
20161028 17:43:58< Soliton> a segfault if you're lucky is much nicer than an exception. of course.
20161028 17:44:03< vultraz> what does that have to do with 'newbie c++'
20161028 17:44:06< celticminstrel> vultraz: -> is dereferencing the pointer.
20161028 17:44:12< celticminstrel> Both * and -> do that.
20161028 17:44:20< vultraz> I see
20161028 17:44:35< celticminstrel> a->b is really just syntactic sugar for (*a)->b
20161028 17:44:38< JyrkiVesterinen> Right now we get neither a segfault nor an exception. We get a log message instead.
20161028 17:44:48< celticminstrel> (Assuming a is a bare pointer and not a class that overloads operator->)
20161028 17:44:53< Soliton> from an exception.
20161028 17:44:54< JyrkiVesterinen> Which is exactly the "defensive programming" way of thinking.
20161028 17:45:14< celticminstrel> How is a segfault nicer than an exception?
20161028 17:47:22< Soliton> an excellent question.
20161028 17:47:33< Soliton> maybe vultraz can elaborate.
20161028 17:48:06< JyrkiVesterinen> Huh, you're right. It is indeed an exception.
20161028 17:48:07< JyrkiVesterinen> https://github.com/wesnoth/wesnoth/blob/master/src/config.cpp#L95-L99
20161028 17:48:43< celticminstrel> That's not the error I was getting though...
20161028 17:48:54< celticminstrel> The error I was getting is the one in check_valid().
20161028 17:50:43-!- Appleman1234 [~Appleman1@KD106181169123.au-net.ne.jp] has quit [Ping timeout: 256 seconds]
20161028 17:51:47< vultraz> how IS a segfault nicer than an exception
20161028 17:52:21< JyrkiVesterinen> Ah. Config::child() is overloaded. Some variants throw an exception if the child doesn't exist, whereas others only log a message and return the null config.
20161028 17:52:23< JyrkiVesterinen> https://github.com/wesnoth/wesnoth/blob/master/src/config.cpp#L659-L669
20161028 17:53:07< JyrkiVesterinen> Responding to the question: a segfault is not nicer than an exception. However, it is nicer than a log message.
20161028 17:53:08< vultraz> ....fffuuunn
20161028 17:53:27< JyrkiVesterinen> For the simple reason that you can ignore a log message, but you can't ignore a segfault.
20161028 17:53:28< celticminstrel> I'm guessing Soliton was being sarcastic, then.
20161028 17:54:39< vultraz> I consider this paradigm-breaking syntax worse than any potential segfaults OR assertions
20161028 17:54:53< vultraz> It's seriously confusing to people reading the code
20161028 17:55:06< celticminstrel> And these functions are unhelpful: https://github.com/wesnoth/wesnoth/blob/master/src/config.cpp#L443-L453
20161028 17:55:17< celticminstrel> That's the error message I was getting in the map generator.
20161028 17:55:42< Soliton> the alternative is a segfault.
20161028 17:55:57< celticminstrel> In the debugger I was able to track it to something in the dependency manager, though that didn't help with figuring out the fix.
20161028 17:56:50< JyrkiVesterinen> Running the game with "--log-debug config" should print a more useful message when the child isn't found.
20161028 17:57:15< celticminstrel> I can try that, I guess.
20161028 17:57:29< celticminstrel> Without a debugger attached I'd need to open Console.app to view the logs...
20161028 17:59:53-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20161028 18:13:11-!- tomreyn_ is now known as tomreyn
20161028 18:18:14< gfgtdf> i think we migth want to add a 'child_or_error' function in the case that the prgorammer _knows_ that the config has such a child. which then thorws an erro directly.
20161028 18:19:26< gfgtdf> then ina  secodn step we can add a s tool that is conditionally compled that checks for all configs that they wre acrually checked before they are used unless required with 'child_or_error'
20161028 18:23:43< JyrkiVesterinen> It turned out that something like child_or_error() already exists, as these two overloads of child().
20161028 18:23:44< JyrkiVesterinen> https://github.com/wesnoth/wesnoth/blob/master/src/config.hpp#L508-L540
20161028 18:25:22< gfgtdf> JyrkiVesterinen: but hey require yout to know the parent tganame
20161028 18:27:27-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20161028 18:28:43-!- irker139 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20161028 18:28:44< irker139> wesnoth: Jyrki Vesterinen wesnoth:fixed-size-chatbox 7ead84efc8ad / / (8 files in 3 dirs): WIP: a widget that forces its child widget to a fixed size https://github.com/wesnoth/wesnoth/commit/7ead84efc8ad8ebed9666988e697f39f9db7bf1b
20161028 18:28:44< irker139> wesnoth: Jyrki Vesterinen wesnoth:fixed-size-chatbox b7170e7e0a93 / / (8 files in 3 dirs): Minor fixes https://github.com/wesnoth/wesnoth/commit/b7170e7e0a93ddcdfd73055e5f8b7a3d3f150bd4
20161028 18:29:00< gfgtdf> JyrkiVesterinen: actuall i now think i heva better idea: we use write a pointer wrapper class liek config_ptr retunred by child(), that asserts false if: 1) The internal values is null or 2) the it wasnt checked for null before beeing used. The second  would require an additional bool variable in it, We could make it so that this only happens when compiled with DEBUG.
20161028 18:30:14< JyrkiVesterinen> Hmm, it sounds like a good idea.
20161028 18:30:29< JyrkiVesterinen> (But don't expect me to create that class, my plate is full.)
20161028 18:30:56< gfgtdf> celticminstrel: the dusdvantage of the pointer (or pointer wrapper) appraoch woudl that we'd need to add quite some (*cfg)["attributebname"] codes whcih aren't that pretty
20161028 18:31:39< JyrkiVesterinen> Ah. That explains why the null config exists.
20161028 18:32:26< JyrkiVesterinen> I thought earlier that implementing the null config is a lot of effort for the sole purpose of giving programmers a reference instead of a pointer.
20161028 18:32:51< JyrkiVesterinen> But indeed, with a reference the *indexing* syntax is much better.
20161028 18:33:16< JyrkiVesterinen> And indexing is a common operation with configs.
20161028 18:33:35-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161028 18:47:07-!- Appleman1234 [~Appleman1@KD106181165216.au-net.ne.jp] has joined #wesnoth-dev
20161028 18:51:42-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20161028 18:51:45-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has quit [Client Quit]
20161028 19:07:02-!- travis-ci [~travis-ci@ec2-54-158-80-32.compute-1.amazonaws.com] has joined #wesnoth-dev
20161028 19:07:03< travis-ci> wesnoth/wesnoth#11795 (fixed-size-chatbox - b7170e7 : Jyrki Vesterinen): The build was fixed.
20161028 19:07:04< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/171465458
20161028 19:07:04-!- travis-ci [~travis-ci@ec2-54-158-80-32.compute-1.amazonaws.com] has left #wesnoth-dev []
20161028 19:15:51-!- Gambit is now known as Gambit_
20161028 19:15:54-!- Gambit_ is now known as Gambit
20161028 19:24:59-!- mjs-de [~mjs-de@x4e30c8e0.dyn.telefonica.de] has quit [Remote host closed the connection]
20161028 19:27:06< irker139> wesnoth: Jyrki Vesterinen wesnoth:fixed-size-chatbox 53af8294b144 / / (8 files in 3 dirs): Minor fixes https://github.com/wesnoth/wesnoth/commit/53af8294b14410ea2988a6b0b27ad89cb1e4ad7f
20161028 19:27:08< irker139> wesnoth: Jyrki Vesterinen wesnoth:fixed-size-chatbox 2637d0302923 / data/gui/ (macros/_initial.cfg schema.cfg widget/size_lock_default.cfg): The WML part https://github.com/wesnoth/wesnoth/commit/2637d03029233836177fea9a2860dc36a104064f
20161028 19:54:02-!- JyrkiVesterinen [~JyrkiVest@87-100-133-159.bb.dnainternet.fi] has quit [Quit: .]
20161028 20:00:51< celticminstrel> [Oct 28@2:30:56pm] gfgtdf: celticminstrel: the dusdvantage of the pointer (or pointer wrapper) appraoch woudl that we'd need to add quite some (*cfg)["attributebname"] codes whcih aren't that pretty
20161028 20:00:51< celticminstrel> Or make the config pointer overload operator[]?
20161028 20:02:28< gfgtdf> celticminstrel: hmm this could be done but often peole assume that pointer[v] is the same as "*(pointer + v)" for pointer types so this might be unexpected.
20161028 20:05:09-!- blackspirit [~blackspir@adsl-128.37.6.49.tellas.gr] has joined #wesnoth-dev
20161028 20:06:36< blackspirit> Hello :)
20161028 20:10:37< gfgtdf> blackspirit: hi
20161028 20:36:31< celticminstrel> Hi
20161028 20:44:00< celticminstrel> ...I just thought. Could a scenario with no sides cause that missing mandatory child error?
20161028 20:44:40< celticminstrel> I didn't have any sides because I wanted to test the main generation part before adding the sides...
20161028 20:44:46< gfgtdf> celticminstrel: thats quite possible.
20161028 20:44:57< celticminstrel> Hmm. Still getting it though. :|
20161028 20:45:18< gfgtdf> celticminstrel: in any case suhc an error is is a bug on the c++ engine.
20161028 20:45:42< celticminstrel> Sure, but I still can't figure out what tag is missing or even where it's missing.
20161028 20:46:10-!- mkdroid [~null@unaffiliated/matthiaskrgr] has joined #wesnoth-dev
20161028 20:46:20< gfgtdf> celticminstrel: do you have a stacktrace of where teh excetion is thorn ?
20161028 20:46:27< celticminstrel> I can get one.
20161028 20:46:27< gfgtdf> thrown*
20161028 20:46:52< gfgtdf> celticminstrel: i think that'd b the fastest way to figue it out the bug.
20161028 20:46:56< celticminstrel> It's something to do with the dependency manager.
20161028 20:47:15< celticminstrel> Shall I post the stack trace for you to take a look?
20161028 20:47:20< gfgtdf> celticminstrel: yes please
20161028 20:48:05< gfgtdf> celticminstrel: does you test scnario use map_generation or scenario_generation ?
20161028 20:48:13< celticminstrel> scenario_generation
20161028 20:48:29< celticminstrel> It was using map_generation earlier though...
20161028 20:48:42< gfgtdf> celticminstrel: so the erro also happen with map_generation ?
20161028 20:48:54< celticminstrel> I think so, let me check...
20161028 20:50:47< celticminstrel> I started by copying the basic random map scenario and then changing it to use the Lua cave generator. I left in all the stuff the cave generator doesn't use, but that can't be the problem since the [generator] tag is just passed verbatim to the generator code.
20161028 20:51:28< celticminstrel> (To the point that it would probably be possible for user configuration to alter the create_map or user_config keys...)
20161028 20:52:12< celticminstrel> (Though that would require my uncommitted work... I don't think the user configuration implementation for Lua generators is fully functional in master.)
20161028 20:52:43< gfgtdf> celticminstrel: did you test with map_generation?
20161028 20:52:55< celticminstrel> Wesnoth is in the process of launching, so not yet.
20161028 20:54:12< celticminstrel> Well, the map actually generates when I use map_generation.
20161028 20:54:25< celticminstrel> Rather than just getting a grassy square.
20161028 20:54:48< celticminstrel> And I don't get the missing child error.3
20161028 20:55:09< blackspirit> what is so special about t_string(s) ?
20161028 20:55:23< celticminstrel> blackspirit: It's a translatable string.
20161028 20:56:12< blackspirit> celticminstrel I have never heard of translatable strings, what do they do ?
20161028 20:56:14< gfgtdf> celticminstrel: note that these mainline mp rnadom scnearios  have the id= both i the inner and teh outer [scenario] https://github.com/wesnoth/wesnoth/blob/master/data/multiplayer/scenarios/Random_Scenario.cfg#L12
20161028 20:56:25< celticminstrel> gfgtdf: I have that also.
20161028 20:56:29< gfgtdf> celticminstrel: hm ok
20161028 20:56:37< gfgtdf> celticminstrel: and its oth exactl the same
20161028 20:56:39< gfgtdf> ?
20161028 20:56:44< gfgtdf> both exactly*
20161028 20:57:51< celticminstrel> gfgtdf: Oh, it considers create_map to be a required key even when using scenario_generation, which seems wrong.
20161028 20:58:27< gfgtdf> celticminstrel: i cannot find create_map= in the scenario í linked above.
20161028 20:58:44< celticminstrel> That would be because it uses the default generator, not the Lua generator.
20161028 20:59:21< gfgtdf> celticminstrel: ah ok
20161028 21:01:33< celticminstrel> Unfortunately, I don't think there's anything I can do about this...
20161028 21:01:36< gfgtdf> celticminstrel: hmm yes sounds wrong.
20161028 21:01:52< celticminstrel> Because the generator has no idea on construction whether it'll be used to generate maps or scenarios.
20161028 21:02:09< gfgtdf> celticminstrel: yes but still you cna rmeove the code that make it manditory
20161028 21:02:13< gfgtdf> celticminstrel: my suggestion is
20161028 21:02:34-!- Kwandulin [~Miranda@p5DDD2B8F.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20161028 21:02:37< gfgtdf> celticminstrel: to make it onyl maditory that at let one of create_scenario or create_map is there
20161028 21:02:47< celticminstrel> That's kinda what I was thinking, too...
20161028 21:02:55< gfgtdf> celticminstrel: so that then if oynl one is given they can 'fallback' to use the other in case the other is required
20161028 21:03:24< gfgtdf> celticminstrel: we already code that handles whta shoudl be done when "create_map" is used with "scenario_generation"
20161028 21:03:50< celticminstrel> That's handled in the map generator base class.
20161028 21:03:52< gfgtdf> celticminstrel: for the oipposite case when "create_scenario" is used with "map_generation" we woudl just generate the scenario and returns its map_data atribute
20161028 21:03:56-!- iwaim [~iwaim@rasteenie.alib.jp] has quit [Ping timeout: 260 seconds]
20161028 21:04:04< celticminstrel> I think I'll handle this one in the Lua generator instead.
20161028 21:04:20< celticminstrel> Well, technically the other is handled in the Lua generator but implemented in the base class.
20161028 21:04:35< celticminstrel> But I don't want to implement the other in the base class because that opens up potential for stack overflow.
20161028 21:05:40< celticminstrel> Admittedly it's a pretty unlikely case...
20161028 21:05:42< gfgtdf> well but here 'potential' means in case that someone adds a 'bad' map_generator subclass
20161028 21:05:46< celticminstrel> Yeah.
20161028 21:06:00< celticminstrel> Hmm, I suppose I could add an implementation of the pure virtual method though.
20161028 21:06:22< celticminstrel> Or wait...
20161028 21:06:24< gfgtdf> celticminstrel: haha
20161028 21:06:34< celticminstrel> Why does the superclass method not take a seed...
20161028 21:06:48< celticminstrel> Wait, that's not the problem.
20161028 21:06:55< celticminstrel> The seed just isn't being passed.
20161028 21:07:05< celticminstrel> gfgtdf: What's funny?
20161028 21:07:22< gfgtdf> celticminstrel: that you found a usecase for such rare feature.
20161028 21:08:28< gfgtdf> celticminstrel: 'not bbeing passed' you mean here https://github.com/wesnoth/wesnoth/blob/master/src/generators/lua_map_generator.cpp#L70 ?
20161028 21:08:44< celticminstrel> Yeah.
20161028 21:08:56< gfgtdf> celticminstrel: hmm shodul be easy to fix.
20161028 21:09:04< celticminstrel> Yeah. I'll fix it.
20161028 21:09:13< celticminstrel> Well, I already did.
20161028 21:09:35-!- timotei [~timotei@wesnoth/developer/timotei] has quit [Read error: Connection reset by peer]
20161028 21:11:37-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20161028 21:11:50-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161028 21:13:12-!- timotei [~timotei@wesnoth/developer/timotei] has joined #wesnoth-dev
20161028 21:14:50-!- iwaim [~iwaim@rasteenie.alib.jp] has joined #wesnoth-dev
20161028 21:16:02< celticminstrel> Now I'm not getting the error anymore...?
20161028 21:16:28< celticminstrel> There's a different error which seems to be getting suppressed though...
20161028 21:18:02< celticminstrel> Ah, it only appears once I click Create Game. Whatever.
20161028 21:19:43< celticminstrel> Looks like it's working now, although I'm pretty sure that chamber is not using the side I told it to...
20161028 21:20:10< celticminstrel> Aha, a typo in the script makes it ignore the size!
20161028 21:20:28< celticminstrel> That's much better!
20161028 21:20:49-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161028 21:22:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161028 21:22:20< celticminstrel> It was also ignoring the jagged value.
20161028 21:23:26< celticminstrel> This makes the HTTT17 map look better too, I think...
20161028 21:25:16-!- mkdroid [~null@unaffiliated/matthiaskrgr] has quit [Ping timeout: 250 seconds]
20161028 21:31:35< celticminstrel> What did passage lazines mean again...
20161028 21:31:38< celticminstrel> ^laziness
20161028 21:37:46< Aginor> morning
20161028 21:38:06< vultraz> morning!
20161028 21:38:49< Aginor> hmm
20161028 21:38:56< Aginor> I should get an ssd :D
20161028 21:39:03< vultraz> everyone should get an ssd
20161028 21:39:43< Aginor> I've been eyeing up the latest i7, 32gb of ram and an ssd or two
20161028 21:40:26< Aginor> but I cannot justify the cost :D
20161028 21:40:31< vultraz> pfft
20161028 21:40:52< zookeeper> well an ssd doesn't cost a lot. unless you want a big one.
20161028 21:41:51< Aginor> I do want a big one
20161028 21:41:57< Aginor> what's the point in a small one?
20161028 21:42:04< Aginor> :D
20161028 21:42:19< Aginor> I also want to underprovision my big one a bit, as that'll extend its lifespan
20161028 21:43:19 * Aginor raises an eyebrow
20161028 21:43:31< Aginor> as you go over 512gb there's a dramatic price jump
20161028 21:43:43< Aginor> although some of the ones I looked at might be enterprise grade
20161028 21:44:08< zookeeper> yeah. i'd be happy with 512gb myself since i apparently just don't need that much space.
20161028 21:47:03< zookeeper> if i'd have 1tb i'd be set for life :p
20161028 21:47:40< Aginor> I'm still keeping a bunch of datasets around
20161028 21:48:02< Aginor> I could probably stop, but...
20161028 21:48:34< irker139> wesnoth: Severin Glöckner wesnoth:lua_mapgen_stuff 1606bb72cef2 / data/core/ (4 files in 4 dirs): Made Drake Flying Animations working on more terrains. https://github.com/wesnoth/wesnoth/commit/1606bb72cef2c2bde02ec3f2227c7b5d138ada91
20161028 21:48:36< irker139> wesnoth: Severin Glöckner wesnoth:lua_mapgen_stuff a8d63e063f8d / data/core/ (macros/animation-utils.cfg units/goblins/Wolf_Rider.cfg units/monsters/Wolf.cfg): reverted addition of ^Vhs https://github.com/wesnoth/wesnoth/commit/a8d63e063f8d31d12a1b3110cd193cf5dfa3179a
20161028 21:48:38< irker139> wesnoth: Charles Dang wesnoth:lua_mapgen_stuff d55265a86c9d / src/gui/widgets/ (slider.cpp slider.hpp): Slider: comment out slider snap-to-value code (e2d3d585be48) https://github.com/wesnoth/wesnoth/commit/d55265a86c9d6f78c84ab51b0241223b0c356db9
20161028 21:48:41< irker139> wesnoth: Severin Glöckner wesnoth:lua_mapgen_stuff 07aea4009188 / data/core/ (terrain.cfg units/goblins/Wolf_Rider.cfg units/monsters/Wolf.cfg): changed submerge to 0.4 in aquatic castles https://github.com/wesnoth/wesnoth/commit/07aea40091888a1a1196444c58a45eb926752f35
20161028 21:48:43< irker139> wesnoth: Lari Nieminen wesnoth:lua_mapgen_stuff 7950cdadd586 / data/core/ (5 files in 5 dirs): Merge pull request #824 from sevu/master https://github.com/wesnoth/wesnoth/commit/7950cdadd5867de5b9b8e9e671326fc37300bde0
20161028 21:48:44< celticminstrel> ...
20161028 21:48:45< irker139> wesnoth: Charles Dang wesnoth:lua_mapgen_stuff a47fb5a9cd62 / src/gui/widgets/ (tree_view_node.cpp tree_view_node.hpp): Tree View Node: implement recursive fold and unfold https://github.com/wesnoth/wesnoth/commit/a47fb5a9cd62a08eba860a8b5b89002ac75e7fe9
20161028 21:48:48< irker139> wesnoth: Charles Dang wesnoth:lua_mapgen_stuff bea53b5ea12f / src/gui/dialogs/gamestate_inspector.cpp: Gamestate Inspector: use recursive fold/unfold instead of widget walker https://github.com/wesnoth/wesnoth/commit/bea53b5ea12fd971cd7eab750289cd4cce71ca22
20161028 21:48:51< irker139> wesnoth: Charles Dang wesnoth:lua_mapgen_stuff b3371b59b449 / src/gui/widgets/tree_view_node.hpp: Tree View Node: use enum instead of integer constants for set_callback_state_cha https://github.com/wesnoth/wesnoth/commit/b3371b59b44908b99aaea4f03808251a3b0d5d7e
20161028 21:48:51 * zookeeper blinks
20161028 21:48:54< irker139> wesnoth: Charles Dang wesnoth:lua_mapgen_stuff a1fe00edacbd / src/gui/dialogs/gamestate_inspector.cpp: Fixup bea53b5ea12f https://github.com/wesnoth/wesnoth/commit/a1fe00edacbde195d381ba2b638601f44f06fa56
20161028 21:48:54< celticminstrel> Wait why are all those extra commits appearing. :(
20161028 21:48:57< irker139> wesnoth: Charles Dang wesnoth:lua_mapgen_stuff 3d64e08520d6 / src/gui/widgets/ (tree_view_node.cpp tree_view_node.hpp): Tree View Node: general cleanup and simplification https://github.com/wesnoth/wesnoth/commit/3d64e08520d6ca4ef98ac163b586264313f0fd94
20161028 21:49:00< irker139> wesnoth: Charles Dang wesnoth:lua_mapgen_stuff 45ae07055f72 / src/gui/dialogs/gamestate_inspector.cpp: Small code tweak as suggested by @jyrkive https://github.com/wesnoth/wesnoth/commit/45ae07055f7202df1f6f40e029724213fa3bb37e
20161028 21:49:03< irker139> wesnoth: Charles Dang wesnoth:lua_mapgen_stuff 0bc568ede559 / src/gui/dialogs/gamestate_inspector.cpp: Gamestate Inspector: disabled recursive folding/unfolding https://github.com/wesnoth/wesnoth/commit/0bc568ede5593f996e26b4f30bbd790881a46869
20161028 21:49:06< irker139> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff 9d9d7450982a / / (8 files in 6 dirs): Merge pull request #845 from wesnoth/lua_mapgen_stuff https://github.com/wesnoth/wesnoth/commit/9d9d7450982aaa70e2d3da85444e568a2b032b31
20161028 21:49:07< celticminstrel> Sorry.
20161028 21:49:08< irker139> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff 51a8cf1f207b / data/lua/cave_map_generator.lua: Lua Cave Mapgen: Support comma-separated chamber locations https://github.com/wesnoth/wesnoth/commit/51a8cf1f207b1c5e7ec0cda6e5d84b9896eb6357
20161028 21:49:10< irker139> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff ff4671809109 / / (5 files in 4 dirs): Lua mapgen: Fix random seed not being used to generate a scenario from a map gen https://github.com/wesnoth/wesnoth/commit/ff46718091094f8fa58f74a54035aba7f4a468dc
20161028 21:49:12-!- mode/#wesnoth-dev [+o shadowm] by ChanServ
20161028 21:49:12< irker139> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff 5dc888e67600 / src/generators/ (lua_map_generator.cpp map_generator.cpp): Lua mapgen: Fix create_map being required for scenario generation https://github.com/wesnoth/wesnoth/commit/5dc888e67600481794be360becd965a9e21b0255
20161028 21:49:14< irker139> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff e79c377b9596 / data/lua/cave_map_generator.lua: Lua mapgen bugfixes: size/jagged ignored for chamber https://github.com/wesnoth/wesnoth/commit/e79c377b9596229d9cecaf26556394d93166374b
20161028 21:49:16< irker139> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff 9c49061a9a51 / data/lua/cave_map_generator.lua: Lua Cave Mapgen: Allow alternate clear and village terrains https://github.com/wesnoth/wesnoth/commit/9c49061a9a516b472bcf0ec35d0627107db6f553
20161028 21:49:18< irker139> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff 3a133db42dca / data/lua/cave_map_generator.lua: Lua Cave Mapgen: Allow a chamber's existence to depend on number of players https://github.com/wesnoth/wesnoth/commit/3a133db42dca9d9331d11a1c1613aa1ac397b07f
20161028 21:49:23-!- mode/#wesnoth-dev [+qz irker*!*@*] by shadowm
20161028 21:49:38<@shadowm> For how much longer is it supposed to -- oh, I guess that's all.
20161028 21:49:40< zookeeper> i would have just about managed to check what the commands were by this point...
20161028 21:49:45< celticminstrel> Yeah, it just finished.
20161028 21:49:57-!- mode/#wesnoth-dev [-qzo irker*!*@* shadowm] by shadowm
20161028 22:02:36-!- atarocch_ [~atarocch@93.56.160.28] has quit [Remote host closed the connection]
20161028 22:06:04-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20161028 22:09:10< mattsc> tad_carlucci: I’ve come across one incompatibility due to the integer/float thing with the new Lua version.
20161028 22:09:43-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection reset by peer]
20161028 22:10:33< mattsc> [gold] used to work with non-integer values.  (I have a scenario in which somebody gives half of previously saved gold back; since it previously worked with non-integers, I did not bother with rounding.  I bet I’m not the only one doing that.)
20161028 22:11:17< mattsc> Now, if the amount is non-integer, it throws a rather cryptic error message.
20161028 22:14:17-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20161028 22:20:36< Aginor> hmm
20161028 22:20:46< mattsc> And just to avoid misunderstandings, the question is not how to fix this, but that it is something that used to work and stopped, but more than that, that it is very hard from the error message to figure out what it going on.
20161028 22:21:05-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20161028 22:21:11< Aginor> I've now done the same blending test as the other night, but recreated in a pure c program that uses SDL
20161028 22:21:21< Aginor> it works/looks as expected here
20161028 22:21:37< mattsc> For reference, the error message: http://pastebin.com/C49faBYC
20161028 22:22:29< mattsc> However, there’s an even bigger issue.  Does this mean that we now need to check every single WML tag implemented in Lua whether it causes trouble when input is non-integer?
20161028 22:23:08< Aginor> mattsc's problem is sounding like a rather bad regression
20161028 22:23:25-!- blackspirit [~blackspir@adsl-128.37.6.49.tellas.gr] has quit [Quit: Leaving]
20161028 22:23:46< mattsc> Aginor: well, it’s trivial to fix for the [gold] tag also.
20161028 22:24:45< Aginor> yes, but if we need to go through and ensure that everything is fine, or this has become an inadvertant API change, it's not ideal
20161028 22:24:52< mattsc> I am more worried about my last question.  Because I only found this by accident.  First time playing through my scenario, gold was even and there was no problem.  Then I went back because something else had not worked and by chance the gold amount was odd.  Otherwise I would not have found it.
20161028 22:25:27< mattsc> Aginor: yes, exactly.  Thinking the same thing there.
20161028 22:25:52< mattsc> Anyways, I need to be off again.  I’ll check back in later.
20161028 22:26:09-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20161028 22:27:31< Aginor> why are we using bitmaps and drawing them for our solidly coloured background?
20161028 22:27:40-!- travis-ci [~travis-ci@ec2-54-81-87-44.compute-1.amazonaws.com] has joined #wesnoth-dev
20161028 22:27:41< travis-ci> wesnoth/wesnoth#11797 (lua_mapgen_stuff - 3a133db : Celtic Minstrel): The build passed.
20161028 22:27:41< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/171520283
20161028 22:27:41-!- travis-ci [~travis-ci@ec2-54-81-87-44.compute-1.amazonaws.com] has left #wesnoth-dev []
20161028 22:27:49< Aginor> as opposed to simply drawing a rectangle of the appropriate colour
20161028 22:30:16< zookeeper> are you sure it's solidly coloured? anyway, maybe it wasn't solidly coloured before
20161028 22:33:30< Aginor> no, squinting at it up close shows that it isn't solidly coloured
20161028 22:33:34< Aginor> so my bad :)
20161028 22:36:07< gfgtdf> celticminstrel: are you doing te this mapgen stuff becaue you intend to use the cavegen in your addon ?
20161028 22:37:33-!- Appleman1234 [~Appleman1@KD106181165216.au-net.ne.jp] has quit [Ping timeout: 265 seconds]
20161028 22:39:45< celticminstrel> I have no plans for that right now. I'd like to get it to a point where the terrain mask in HTTT17 can be removed though.
20161028 22:40:39-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit []
20161028 22:45:24< gfgtdf> celticminstrel: i wonder whether theat mask is just random terrain or something intentional, 
20161028 22:46:08< celticminstrel> I have no idea, because I can't load maps in the editor.
20161028 22:46:23< celticminstrel> However, I doubt it's anything other than random terrain.
20161028 22:47:33< gfgtdf> zookeeper:  maybe you know^?
20161028 23:03:13-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Quit: ChipmunkV]
20161028 23:03:28-!- Appleman1234 [~Appleman1@KD036012040083.au-net.ne.jp] has joined #wesnoth-dev
20161028 23:23:15-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161028 23:35:34-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20161028 23:44:12-!- gfgtdf [~chatzilla@x4e3638a9.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 49.0.2/20161019084923]]
20161028 23:49:03-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
--- Log closed Sat Oct 29 00:00:36 2016