--- Log opened Tue Sep 26 00:00:10 2017
20170926 01:01:19-!- vultraz [uid24821@wesnoth/developer/vultraz] has quit [Quit: Connection closed for inactivity]
20170926 01:33:26-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20170926 02:06:41-!- vultraz [uid24821@wesnoth/developer/vultraz] has joined #wesnoth-dev
20170926 02:12:09-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20170926 02:17:07-!- JyrkiVesterinen [~JyrkiVest@85-23-197-3.bb.dnainternet.fi] has joined #wesnoth-dev
20170926 02:30:52< Necrosporus> scons is trying to put all *.o files in libwesnoth_extras.a and fails because it doesn't have enough space on device
20170926 02:31:09< Necrosporus> Why does it have to ar all debug *.o files? cmake doesn't do that
20170926 03:09:59-!- irker343 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20170926 03:51:12-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
20170926 03:55:56-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20170926 04:01:45-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20170926 04:01:57-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Client Quit]
20170926 04:40:12-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20170926 04:40:18-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20170926 05:08:32-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20170926 05:23:57< vultraz> out of idle curiosity, how much work would it be to allow actual videos to be played in wesnoth? It seems it would be a  lot more efficient than fast drawing of jpgs like TLU is apparently doing
20170926 05:26:16< vultraz> Would probably need to incorporate ffmpeg or something
20170926 05:33:25-!- JyrkiVesterinen [~JyrkiVest@85-23-197-3.bb.dnainternet.fi] has quit [Quit: Going to work]
20170926 05:45:17-!- Kwandulin [~Kwandulin@pD9FD50F1.dip0.t-ipconnect.de] has joined #wesnoth-dev
20170926 05:55:19< Necrosporus> Probably mplayer
20170926 05:55:28< Necrosporus> But it should not be enabled by default 
20170926 05:55:55< Necrosporus> at least not required
20170926 05:56:23< Necrosporus> Though do you really want all existing formats supported, perhaps it's better to limit support to one video format such as webm?
20170926 05:59:21< vn971> Necrosporus: talking about formats, is webm really free (as code and from patents)? I thought ogg is the free video flagman.
20170926 05:59:38< Necrosporus> I think it is
20170926 05:59:47< Necrosporus> ogg is a container format
20170926 06:00:11< vn971> at least I see ogg in free software, never webm. ogg+ogv I mean.
20170926 06:00:11< Necrosporus> It does usually contain Vorbis or Opus for audio and Theora for video
20170926 06:00:21< vn971> Yup, as you wrote.
20170926 06:00:26< Necrosporus> Don't say 'ogg' since it's just a container
20170926 06:00:33< vn971> ok-ok, fair.
20170926 06:00:55< Necrosporus> I think Theora is outdated for now
20170926 06:01:11< Necrosporus> there are more efficient video compressors available
20170926 06:01:35< Necrosporus> If we are to pick one perhaps it should be patent-free and more or less efficient
20170926 06:23:47-!- JyrkiVesterinen [~JyrkiVest@195-192-251-124.s1networks.fi] has joined #wesnoth-dev
20170926 06:33:19< JyrkiVesterinen> I think VP9 would be a good video format.
20170926 06:34:01< JyrkiVesterinen> It's not patent-free, but Google grants a good patent license: https://www.webmproject.org/license/bitstream/
20170926 06:34:25< JyrkiVesterinen> VP9 is efficient, and well battle-tested, as it's YouTube's default video format.
20170926 06:39:54-!- irker800 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20170926 06:39:54< irker800> wesnoth: Wedge009 wesnoth:master 3df86a0f6cae / doc/man/wesnoth.6: Spelling corrections for 19b1f0a6 and 9b7db054. https://github.com/wesnoth/wesnoth/commit/3df86a0f6cae6a924c148cc2179a320b6d5c6868
20170926 06:58:28< Necrosporus> JyrkiVesterinen, though the best streaming format is not necessarily best for local files
20170926 06:58:47< Necrosporus> not necessarily worst either
20170926 06:59:23< JyrkiVesterinen> Needs for streaming and local files are similar enough that I think being a good streaming format is a good indicator.
20170926 07:00:11< wedge009> I just noticed misc/found_poetry/. How bizarre - it's not as though the dA account is dead.
20170926 07:15:53-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20170926 07:25:24-!- atarocch [~atarocch@93.56.164.28] has joined #wesnoth-dev
20170926 07:32:26-!- JyrkiVesterinen_ [~JyrkiVest@195-192-251-124.s1networks.fi] has joined #wesnoth-dev
20170926 07:34:55-!- JyrkiVesterinen [~JyrkiVest@195-192-251-124.s1networks.fi] has quit [Ping timeout: 248 seconds]
20170926 07:35:01-!- JyrkiVesterinen_ is now known as JyrkiVesterinen
20170926 08:26:28-!- vultraz [uid24821@wesnoth/developer/vultraz] has quit [Quit: Connection closed for inactivity]
20170926 08:45:05-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20170926 09:21:54-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170926 09:40:31-!- irker800 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20170926 10:05:55-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170926 10:06:29-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170926 10:15:48-!- Kwandulin [~Kwandulin@pD9FD50F1.dip0.t-ipconnect.de] has quit [Ping timeout: 252 seconds]
20170926 10:17:04-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20170926 10:31:53-!- JyrkiVesterinen [~JyrkiVest@195-192-251-124.s1networks.fi] has quit [Quit: Going offline, see ya! (www.adiirc.com)]
20170926 10:44:24-!- JyrkiVesterinen [~JyrkiVest@195-192-251-124.s1networks.fi] has joined #wesnoth-dev
20170926 10:51:08< AI0867> 17:28 < vultraz> that is a huge merge commit ← it is only 'huge' because the pot-update is huge
20170926 10:51:22< AI0867> 17:31 < Necrosporus> It was like if merge commit didn't change anything ← that's because it doesn't
20170926 10:53:07< AI0867> 17:30 < vultraz> please try to avoid merge commits (by rebasing before pushing) in the future ← in this case I agree, but when there are multiple related commits, I do not. rebasing throws away history that may be useful
20170926 11:02:44-!- gfgtdf [~chatzilla@x4e3681d2.dyn.telefonica.de] has joined #wesnoth-dev
20170926 11:03:06< gfgtdf> how so? rebasig usually doesn't merge/remove any commits
20170926 11:13:16< JyrkiVesterinen> I think his point is that rebasing hides the fact that the commits originated form a different branch in the first place.
20170926 11:13:19< JyrkiVesterinen> *from
20170926 11:16:54< gfgtdf> JyrkiVesterinen: that fact is hidden in the way he did it too, i mean ther is no merge commit workingbranch->master only the merge commit master->workingbranch
20170926 11:18:32< JyrkiVesterinen> Right. One should avoid getting "merge master to working branch" commits to master, they are just noise.
20170926 11:19:32< gfgtdf> vultraz: i agree that video sequences coudl be nice to have 
20170926 11:53:04< DeFender1031> I'm not sure I like the idea of video sequences, TBH.
20170926 11:53:24< DeFender1031> I think there's something nice about the sort of simple feel to the game as it is.
20170926 11:54:14< DeFender1031> And I think that adding video sequences would greatly raise the bar for what would be considered "good" content.
20170926 11:55:15< DeFender1031> Or rather, raise the bar in such a way that your average content creator would not be able to keep up with.
20170926 11:55:49< DeFender1031> (And bear in mind that this is ME saying this, and I'm usually all FOR things that make cinematics easier.)
20170926 11:56:03< JyrkiVesterinen> I think it's not a problem.
20170926 11:56:30< JyrkiVesterinen> UMC creators would still be able to not hit the bar. They don't *have* to make content that compares favorably to the best campaigns.
20170926 11:57:33< DeFender1031> I'm not sure I follow.
20170926 11:58:03< zookeeper> what bar? 99% of mainline content wouldn't have video content in the first place.
20170926 11:58:54< DeFender1031> zookeeper, not immediately, no, but down the line?
20170926 11:59:23< zookeeper> if that's the question then the answer is "no"
20170926 11:59:33< DeFender1031> Huh?
20170926 11:59:37< zookeeper> what?
20170926 11:59:47< DeFender1031> If what's the question?
20170926 12:00:02< zookeeper> the thing right above what i said that contains a question mark, naturally
20170926 12:00:03< zookeeper> what else?
20170926 12:00:28< DeFender1031> So you don't think that if there's the capability for video, that mainline would have video added to it eventually?
20170926 12:00:43< DeFender1031> Why do you think that's so unlikely?
20170926 12:01:26< zookeeper> because that's exactly the kind of thing that no one would actually do, at a quality that wouldn't look terrible, anyway
20170926 12:01:44< DeFender1031> hmm...
20170926 12:01:49< DeFender1031> then why add it?
20170926 12:01:53< zookeeper> although i don't know what kinds of usecases you're thinking of.
20170926 12:02:50< zookeeper> the line between "video" and our current animation systems is kind of blurry. if you just mean directly supporting a video format wherever static images are currently used, then sure that would get used in some form.
20170926 12:03:37< JyrkiVesterinen> Yesterday it was pointed out that To Lands Unknown already has a video sequence.
20170926 12:03:41< DeFender1031> zookeeper, honestly, if wesnoth gets as big as everyone here is hoping it is, it wouldn't entirely surprise me if some high-quality larp group decided to make a bunch of live-action cutscenes for it... Granted, there are a LOT of ifs in there, but still.
20170926 12:03:59< JyrkiVesterinen> Just implemented in a horribly inefficient way because the engine doesn't have video support.
20170926 12:04:11< zookeeper> but if you mean video as in cinematic cutscenes or whatever, then those would either 1) look terrible, 2) be created based on the in-game graphics and thus become outdated immediately or 3) just never happen
20170926 12:04:54< zookeeper> DeFender1031, you think wesnoth larp videos wouldn't look incredibly dorky?
20170926 12:05:11< AI0867> 13:16 < gfgtdf> JyrkiVesterinen: that fact is hidden in the way he did it too, i mean ther is no merge commit workingbranch->master only the merge commit  master->workingbranch ← it still shows what the parent commit was. In some cases (not in this case), even that may be useful information
20170926 12:05:33< DeFender1031> JyrkiVesterinen, well, what's the definition of "video sequence" anyway? There are add-ons that have fancy title screen animations. Or my mage fade-ins. Or in-game scripted cutscenes. Or more elaborate animations. What reaches the level of "video sequence"?
20170926 12:05:44< Necrosporus> zookeeper, perhaps some UMC authors will want to add video
20170926 12:05:54< Necrosporus> At least TLU would use it immediately I guess
20170926 12:06:02< AI0867> a couple years back someone actually made a mostly-functional video player in WML
20170926 12:06:15< AI0867> that is, a sequence of images + soundtrack
20170926 12:06:15< Necrosporus> What? With real video format?
20170926 12:06:21< Necrosporus> ah, that's TLU
20170926 12:06:24< AI0867> the problem was synchronising the two
20170926 12:06:34< JyrkiVesterinen> DeFender1031: A "video sequence" is quite simply something that can't be shown in any other way than displaying a large number of images in rapid succession.
20170926 12:06:37< AI0867> before TLU
20170926 12:06:54< DeFender1031> zookeeper, I think they would have the potential to look incredibly dorky, certainly, but I've also seen a LOT of so-called "amateur" film (and really, amateur productions in general) that rival or even surpass some of their big-budget counterparts.
20170926 12:10:12< DeFender1031> But like I said, there are a lot of ifs. That's one of them.
20170926 12:10:13< zookeeper> anyway, i don't of course in principle object to things like supporting a proper video format directly. i just think that uses which actually fit wesnoth are pretty rare and there wouldn't be any particularly extravagant usecases for it in mainline content.
20170926 12:11:52< DeFender1031> Actually, what you said about it being available in any context where images can be used made me reconsider. For small stuff, like a complex titlescreen animation, or other simple animations that can't be easily done with sprite stuff, it might be useful.
20170926 12:12:29< DeFender1031> (Also, the ability to add synched sound to story screen stuff would be useful.)
20170926 12:13:08< zookeeper> yes, certainly. that's the kind of thing i'd happily use it for, although whether it'd work well for pixel artsy animations, i don't know.
20170926 12:13:54< zookeeper> i mean, everything can be done as an image sequence, but in practise it's awkward to have 200 files for one fancy animation.
20170926 12:14:18< JyrkiVesterinen> And wasteful for disk space and memory usage.
20170926 12:14:43< zookeeper> it's just that whenever people say "video sequences", i'm going to assume they're probably thinking of cutscenes of some kind.
20170926 12:15:01< zookeeper> JyrkiVesterinen, that too, i suppose
20170926 12:25:52< DeFender1031> zookeeper, that was my initial thinking as well, which is why I was opposed.
20170926 12:26:17< DeFender1031> If it's mostly going to be used to enhance small parts of the already existing style, that's different.
20170926 12:26:47< DeFender1031> Also, I'd recommend supporting at least one lossless format if that's the case
20170926 12:27:13< DeFender1031> (Since stuff based on pixel stuff will look terrible with compression artifacts.)
20170926 12:30:22< JyrkiVesterinen> Found this Stack Overflow question about lossless video formats: https://stackoverflow.com/q/768999
20170926 12:33:40< DeFender1031> JyrkiVesterinen, this is very old. There are a lot of much better lossless formats that have come out in the last few years.
20170926 12:40:35-!- Kwandulin [~Kwandulin@pD9FD50F1.dip0.t-ipconnect.de] has joined #wesnoth-dev
20170926 12:53:33< DeFender1031> JyrkiVesterinen, H.264 lossless might be better, for example.
20170926 12:56:04< Necrosporus> It is mentioned there as well
20170926 12:57:24-!- Kwandulin [~Kwandulin@pD9FD50F1.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20170926 13:00:06< DeFender1031> Necrosporus, legit.
20170926 13:01:28< Necrosporus> Though it has patent problems unlike say Theora
20170926 13:07:06< DeFender1031> Does it? As far as I understand, it's used pretty freely by anyone who wants to.
20170926 13:07:50< JyrkiVesterinen> https://en.wikipedia.org/wiki/H.264/MPEG-4_AVC#Licensing
20170926 13:08:28< JyrkiVesterinen> In particular, H.264 video decoding requires royalties.
20170926 13:09:10< JyrkiVesterinen> Some software (such as Firefox) have worked around that by using decoders already available in the operating system.
20170926 13:14:40< DeFender1031> Interestring.
20170926 13:16:25< JyrkiVesterinen> I develop an open source music visualizer (Windows-only), and it uses a similar trick. It decodes music with Media Foundation (which comes with Windows), and can therefore support proprietary audio formats such as AAC.
20170926 13:16:42< JyrkiVesterinen> (In fact, I'm listening AAC-encoded music with it right now.)
20170926 13:19:43< DeFender1031> So how do all these things include support on linux?
20170926 13:20:10< JyrkiVesterinen> Via ffmpeg, which you can usually install via repositories.
20170926 13:20:31< DeFender1031> and ffmpeg pays royalties?
20170926 13:20:58< JyrkiVesterinen> I haven't looked at the situation about that.
20170926 13:21:33< DeFender1031> IIRC, though, you can't patent a foram, you can only patent code development techniques for decoding a format, no? If that's correct, then why are there not other libraries for decoding out there that don't violate the patent?
20170926 13:21:41< DeFender1031> format*
20170926 13:24:51< JyrkiVesterinen> Sorry, I don't know the answer to that. I'm not exactly a patent lawyer.
20170926 13:27:01< DeFender1031> Sorry, I was basically just thinking out loud.
20170926 13:27:09< DeFender1031> Sorry.
20170926 13:34:38< Necrosporus> DeFender1031, I think that patent problems arise only if you distrubite it for USA citizen
20170926 13:34:44< Necrosporus> perhaps for some other countries
20170926 13:34:51< Necrosporus> or you are one yourself
20170926 13:35:18< Necrosporus> So distributions such as Fedora which host in USA do not include support for proprietary formats by default
20170926 13:35:47< Necrosporus> Other distributions either host in other countries, for example Ubuntu is from SAR
20170926 13:35:59< DeFender1031> Is wesnoth hosted in USA?
20170926 13:36:05< DeFender1031> I had thought it wasn't.
20170926 13:36:05< Necrosporus> github...
20170926 13:36:10< DeFender1031> Ah, right.
20170926 13:36:28< Necrosporus> Either way it is undesirable for wesnoth to include proprietary codecs
20170926 13:36:45< Necrosporus> It will mean that it won't be able to come into Fedora, Suse and some other's repos
20170926 13:37:02< Necrosporus> Not into Ubuntu main too
20170926 13:37:28< JyrkiVesterinen> At least for lossy video, we should definitely go with a free format, like aforementioned VP9.
20170926 13:37:38< Necrosporus> or vp10?
20170926 13:37:59< Necrosporus> There is also new video format from Xiph.org
20170926 13:38:07< JyrkiVesterinen> https://en.wikipedia.org/wiki/VP9#Successor:_from_VP10_to_AV1
20170926 13:38:16< JyrkiVesterinen> "Accordingly, Google has stated that they will not deploy VP10 internally or officially release it, making VP9 the last of the VPx-based codecs to be released by Google."
20170926 13:38:41< JyrkiVesterinen> Daala is in too early state to be used right now.
20170926 13:38:54< Necrosporus> what about AV1?
20170926 13:39:44< Necrosporus> So, ng formats include Daala, AV1, HEVC, H.265, maybe something else too?
20170926 13:39:58< Necrosporus> We can't use HEVC and H265
20170926 13:40:14< JyrkiVesterinen> Hmm, Wikipedia says that the AV1 bitstream format is projected to be frozen in Q4 2017.
20170926 13:40:49< JyrkiVesterinen> Assuming that someone here would bother to start implementing video support to begin with, it's plausible that AV1 would be an option by then.
20170926 13:41:17< DeFender1031> It'd also be nice if there were some video format that supported both partial and full transparency... but that's not really a thing that exists.
20170926 13:42:00< Necrosporus> Any format could support it though
20170926 13:42:04< DeFender1031> I don't even think any suppot single-bit transparency.
20170926 13:42:20< Necrosporus> You just need a 4 channel format
20170926 13:42:26< Necrosporus> Are there such formats yet?
20170926 13:47:03< DeFender1031> I doubt it. I just checked with a friend of mine who knows formats and he suggested ffv1
20170926 13:49:45< Necrosporus> https://forums.wesnoth.org/viewtopic.php?f=21&t=36057
20170926 13:49:54< Necrosporus> I cannot find 1.12 -> 1.13 thread
20170926 13:55:26-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20170926 14:06:18< gfgtdf> vultraz: do you think we shodul make a sticky about the 1.12 -> 1.13 changes?
20170926 14:12:17-!- JyrkiVesterinen [~JyrkiVest@195-192-251-124.s1networks.fi] has quit [Quit: .]
20170926 14:36:19-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170926 15:07:38-!- Kwandulin [~Kwandulin@pD9FD50F1.dip0.t-ipconnect.de] has joined #wesnoth-dev
20170926 15:29:29< zookeeper> fill in the existing thread in moderator's forum and move it to WML workshop.
20170926 15:29:41< zookeeper> (not necessarily in that order)
20170926 16:28:54< shadowm> Can somebody tell vultraz to ask before assigning issues to people?
20170926 16:29:49< shadowm> I'm just going to write the server side of 2043 and leave him to deal with the client side (technically, it can be a pure server-side change but then the presentation won't be particularly convenient for users).
20170926 16:31:56< Necrosporus> shadowm, are you talking about recent report about confusing server message?
20170926 16:32:04< shadowm> I'm talking about #2043.
20170926 16:32:13< shadowm> You decide what it's about, I'm just the messenger.
20170926 16:32:29< Necrosporus> That's what I mean
20170926 16:32:42< shadowm> Then why are you asking?
20170926 16:33:13< Necrosporus> I wonder what do you think about the problem
20170926 16:33:39< shadowm> I think stuff. Enough said.
20170926 16:34:01< shadowm> In fact I already said what I think, above.
20170926 16:34:30< Necrosporus> Does server only reject space (\x20) ~ \ / and : in filename?
20170926 16:34:45< shadowm> Why do you want to know?
20170926 16:34:47< Necrosporus> Also how is / even possible in filename?
20170926 16:34:51< shadowm> Are you going to write the patch for me? :D
20170926 16:35:25< Necrosporus> I'm curious if you can actually upload addon with filenames full of control characters including \n and \r and ...
20170926 16:35:26-!- vultraz [uid24821@wesnoth/developer/vultraz] has joined #wesnoth-dev
20170926 16:35:46< shadowm> YOu can check the source.
20170926 16:36:12< shadowm> Either that or try it out for yourself. I know the answer, but I don't want to rob you of the wonderful experience of experimenting and tinkering with stuff.
20170926 16:41:22< Necrosporus> if(name.empty() || name == "." || name.find_first_of("/:\\~ \r\n\v\t") != std::string::npos || name.find("..") != std::string::npos)
20170926 16:42:10< Necrosporus> src/addon/validation.cpp:62
20170926 16:42:33< shadowm> Congratulations.
20170926 16:42:42< Necrosporus> shadowm, actually, somehow idea that I can access source code of addon server didn't occur to me
20170926 16:43:58< Necrosporus> Though filename can't possibly contain /, can it?
20170926 16:46:00< Necrosporus> Shouldn't all other control characters in range 0 to 31 and 127 be prohibited too
20170926 16:46:31< Necrosporus> And non-ascii characters can cause problems perhaps
20170926 16:46:40< shadowm> File a patch if you want idc.
20170926 16:46:52< Necrosporus> idc?
20170926 16:47:58< Necrosporus> I'm not sure if they can or can't cause problems
20170926 16:48:22< shadowm> Only insofar that it's a royal PITA to manipulate files with those characters in their names.
20170926 16:49:00< shadowm> (And Windows and macOS might consider them invalid but don't quote me on that.)
20170926 16:51:25-!- JyrkiVesterinen [~JyrkiVest@85-23-197-3.bb.dnainternet.fi] has joined #wesnoth-dev
20170926 16:52:12< Necrosporus> shadowm, I mean what about characters which are not filtered out but still are special or control characters 
20170926 16:52:18< shadowm> YES I KNOW.
20170926 16:54:34< JyrkiVesterinen> Windows indeed disallows control characters in filenames.
20170926 16:58:35< DeFender1031> Necrosporus, "idc" means "I don't care". I'd also suggest that it seems like shadowm is a little stressed at the moment and it might therefore not be a good time to ask a whole bunch of questions.
20170926 17:00:17< Necrosporus> I guess that's right
20170926 17:00:29< Necrosporus> caps also suggest that
20170926 17:01:07< Necrosporus> JyrkiVesterinen, though \127 is not prohibited seems
20170926 17:01:22< Necrosporus> At least MS documentation doesn't list it as disallowed
20170926 17:07:37< JyrkiVesterinen> Indeed, \127 seems to be allowed on Windows.
20170926 17:09:13< DeFender1031> In general when dealing with filenames that will go on systems that I don't have control over, I tend to take a very conservative approach and allow only alphanumeric plus tilde, dash, underscore, and period.
20170926 17:09:32< DeFender1031> IIRC, FAT32 for example will choke on anything else.
20170926 17:09:40< DeFender1031> Oh, and regular spaces, of course.
20170926 17:11:21< shadowm> Sounds like you found yourself a job to do then. :p
20170926 17:11:22< JyrkiVesterinen> There are more characters that FAT32 allows, such as parentheses and exclamation marks.
20170926 17:11:29< DeFender1031> besides, there's enough variety in all of that that add-on creators shouldn't need to be using other wacky names.
20170926 17:12:28< DeFender1031> JyrkiVesterinen, ah, good point.
20170926 17:13:04< DeFender1031> I was just rattling off the top of my head. The point was to disallow any characters that are not accepted by all filesystems.
20170926 17:13:17< DeFender1031> shadowm, you mean for Necrosporus to do. :P
20170926 17:13:32< JyrkiVesterinen> For that, this table is useful: https://en.wikipedia.org/wiki/Filename#Comparison_of_filename_limitations
20170926 17:13:59< shadowm> DeFender1031: He strikes me as more of a philosopher than a coder.
20170926 17:16:01< shadowm> vultraz: Where do we display server errors when an add-on upload fails?
20170926 17:16:22< shadowm> *From* where?
20170926 17:16:36< vultraz>  One sec
20170926 17:16:51< vn971> I wonder. Guys, can we do something about "cross-platform-ness" of wesnoth add-ons? 2 example bugs I know of:
20170926 17:16:54< shadowm> Also, since you can't read logs: <shadowm> Can somebody tell vultraz to ask before assigning issues to people?
20170926 17:16:55< Necrosporus> DeFender1031, remove tilda from your list
20170926 17:17:02< DeFender1031> Necrosporus, why?
20170926 17:17:22< vn971> A. you can create an add-on which would access files in case-ignoring way, and it'll work on Windows. It won't on Mac, Linux and android though.
20170926 17:18:05< shadowm> I thought macOS' default filesystem was case-insensitive case-preserving just like Windows'.
20170926 17:18:12< vultraz> shadowm: https://github.com/wesnoth/wesnoth/blob/master/src/gui/dialogs/addon/manager.cpp#L686-L688
20170926 17:18:13< vn971> B. you can use \n, \r and similar symbols for end of line. I didn't test really, but I expect it to work a bit differently on different platforms, too.
20170926 17:18:26< Necrosporus> shadowm, campaign_server.cpp ?
20170926 17:18:30< shadowm> No Necrosporus.
20170926 17:18:41< shadowm> vn971: Wesnoth can read both CR-LF and LF files on any platform.
20170926 17:19:33< vn971> shadowm: you're right about MacOs. But there's still difference between Linux/Windows.
20170926 17:19:36< shadowm> I thought ages ago about making Wesnoth go out of its way to check the case of preprocessor includes on Windows but I never got around to it.
20170926 17:19:54< shadowm> (Or on any platform, really. Linux can also do case-insensitive filesystems.)
20170926 17:20:08< shadowm> (Conversely, Windows can do case-sensitive filesystems.)
20170926 17:20:11< Necrosporus> I think it should be case-sensitive on all platformss
20170926 17:20:23< vn971> shadowm: yup, that's exactly my point. We could make it fail fast under Windows, which would force add-on developers to keep the case correct.
20170926 17:20:48< Necrosporus> vn971, I'd suggest make wesnoth error about missing image/sound on windows if the case of name doesn't match in WML and in filesystem
20170926 17:21:02< Necrosporus> even if it is techincally accessible
20170926 17:21:06< vn971> About CR-LF and LF. I definitely remember wesnoth NOT showing correct text when I used `\r`.
20170926 17:21:19< Necrosporus> If it doesn't already
20170926 17:21:37< Necrosporus> IT will help windows UMC developers notice they mistake before they post it to server
20170926 17:21:44< vn971> Necrosporus: I suggest the same. So, what would be the proper action then? Raise a bug, or?..
20170926 17:21:50< shadowm> vn971: That's just a carriage return.
20170926 17:22:08< Necrosporus> vn971, first check if it doesn't do it already
20170926 17:22:17< Necrosporus> then if bugreport exists
20170926 17:22:25< shadowm> MacOS 9 and older are the only operating systems where CR was used as a line-ending character on its own as far as I know.
20170926 17:22:38< vn971> Necrosporus: oh, but I don't have Windows..
20170926 17:22:51< Necrosporus> Ask somebody on windows to test
20170926 17:23:05< Necrosporus> So windows allows: ! # $ % & ' ( ) + , - . 
20170926 17:23:05< Necrosporus> ; = @ [ \ ] ^ _ ` { } ~ \127
20170926 17:23:11< vn971> shadowm: OK, if at least both current major systems are supported, no problem, I guess.
20170926 17:23:14< shadowm> I can tell you that if Wesnoth didn't treat CR-LF and LF as the same, most add-ons wouldn't work on the other platform.
20170926 17:23:47< shadowm> I can also attest to myriads of Windows users being able to play the most popular add-on ever, which was written on Linux and therefore is full of solitary LFs.
20170926 17:24:12< DeFender1031> Necrosporus, your list there includes tilde.
20170926 17:24:28< DeFender1031> And you didn't answer my question.
20170926 17:24:33< Necrosporus> It is list of special characters windows allow
20170926 17:25:27< shadowm> That's a lie, by the way. Windows allos : and $ as well.
20170926 17:25:34< shadowm> It just wants you to believe otherwise.
20170926 17:25:54< DeFender1031> shadowm, Necrosporus did mention $
20170926 17:26:04< shadowm> In particular, : crops up in the file names of alternate streams.
20170926 17:26:23< Necrosporus> I think only +-._ could be more or less safely used in filenames out of this
20170926 17:26:41< DeFender1031> I know for a fact that FAT32 protests if you try to use :
20170926 17:27:07< JyrkiVesterinen> Necrosporus: what's the problem with other symbols?
20170926 17:27:45< vn971> Can somebody test Windos case behavior? I've uploaded "Gold Per Village" plugin. If it will have a "gold" icon as its main icon, windows ignores case. If there would be no icon, Windows behaves as Linux (considers the case and refuses to load with case mismatch).
20170926 17:28:10< shadowm> Windows ignores case by default.
20170926 17:28:31< shadowm> I already said this.
20170926 17:28:47< vn971> shadowm: yes. But there was a suggestion that maybe wesnoth already enforces add-on developers to write correct case. Which is exactly I'd like to test.
20170926 17:28:48< Necrosporus> ! is used for shell history expansion, # is comment, $ is variable substition in many languages, % is variable substition in windows, & puts command in background ' is often used for quoting and thus very hard to be quoted
20170926 17:29:17< Necrosporus> ( ) are used for command substition such as $( command ), same goes for `
20170926 17:29:26< shadowm> ALl I can say about that is that the Windows shell places additional restrictions on filenames that the Windows API does not.
20170926 17:29:31< JyrkiVesterinen> Other than ', I don't really see the others as problematic in *file names*.
20170926 17:29:53< Necrosporus> They are problematic if they are shared with unix system
20170926 17:29:55< vn971> shadowm: y, true about column support. I remember windows7 somehow couldn't _delete_ files with ":". So the filesystem supports it, but support from inside win is messy.
20170926 17:29:59< shadowm> First and foremost, colons in filenames can and do happen in NTFS as they are used by a core feature of it.
20170926 17:30:28< vn971> ok, in other words. Anyone with windows who can help me a bit understand the current status? About add-ons and case-sensitivity?
20170926 17:30:32< JyrkiVesterinen> Still, I don't see why it would be a problem if your Unix system has files with, say # characters.
20170926 17:30:34< shadowm> This has been used time and again to hide malware from graphical tools that use the same API as the Windows shell to enumerate files.
20170926 17:30:50< JyrkiVesterinen> If you need to, say, remove such a file, "rm #" works like you'd expect.
20170926 17:30:54< vn971> I ask to go to 1.12 add-ons and check if there would be an icon near "Gold Per Village" add-on.
20170926 17:31:49< vn971> JyrkiVesterinen: no, it won't. :)  But rm '#' will.
20170926 17:31:51< shadowm> vn971: Okay, I'm going to phrase it this way: I was around until 1.13.2 and until then Wesnoth made no efforts at all to check name case on Windows.
20170926 17:32:01< shadowm> That I am 100% certain of.
20170926 17:32:12< shadowm> I'm 99% certain that nobody's done any effort since then to change the status quo.
20170926 17:32:23< vn971> shadowm: ah, understood.
20170926 17:32:24< JyrkiVesterinen> shadowm: Indeed, NTFS has almost zero restrictions in file names. It's the Win32 subsystem that implements the restrictions.
20170926 17:32:38< shadowm> And in particular, this isn't going to change on 1.12.x.
20170926 17:32:41< Necrosporus> , is used to separate files in WML animations, ; is separator too, = is used to delimit variable/key from value, @ is used to delimit username and hostname, [] are used for multichar match such as [a-z], ^ is used as quoting char in windows shell, {} are used for macros, \128 is impossible to display, ~ denotes home directory or ~user directory
20170926 17:32:57< vn971> So maybe we should change this, as Necrosporus and I proposed? It shouldn't be hard, I guess.. And it will certainly help write correct add-ons, won't it?
20170926 17:33:08< Necrosporus> Which leaves only +-._ from the set
20170926 17:33:33< JyrkiVesterinen> As I said, I don't really agree with Necrosporus's reasoning to ban almost all ASCII symbols.
20170926 17:33:37< Soliton> can you even figure out the "real" case on case insensitive filesystems? and can the user fix the case easily (without removing and recreating)?
20170926 17:34:00< JyrkiVesterinen> All case-insensitive filesystems in use today are case-preserving.
20170926 17:34:01< shadowm> You can, as the Windows shell evidently can.
20170926 17:34:09< JyrkiVesterinen> So yes, you know what the correct case is.
20170926 17:34:21< Soliton> and can you rename a file with just changing the case?
20170926 17:34:37< vn971> Soliton: sure. At least WinXP could, the last win I used.:D
20170926 17:34:46< JyrkiVesterinen> No, Windows Explorer doesn't allow you to do that.
20170926 17:35:01< JyrkiVesterinen> A workaround is to change both the case and something else, then revert the other change.
20170926 17:36:41< DeFender1031> idiotic that it doesn't allow that
20170926 17:37:19< DeFender1031> anyway, since wesnoth is not doing things via a shell, Necrosporus's reasoning for excluding a bunch of stuff is not really valid.
20170926 17:37:42< shadowm> The thing is, back when it was first written, the filesystem it targeted was FAT16.
20170926 17:37:46< DeFender1031> (As well, since shells have escape characters, pretty much any filename CAN still be represented in a shell.)
20170926 17:37:56< JyrkiVesterinen> It was quite annoying at one point in my previous job. I needed to rename over a hundred sprites we had gotten from outsourced artists. Most of them had otherwise correct names but with wrong case. :/
20170926 17:38:08< vn971> DeFender1031: there were 2 suggestions. Excluding characters and case-sensitivity. I'd still vote for enabling case-sensitivity, if we can..
20170926 17:38:11< shadowm> I believe even with LFN, it wasn't case-preserving, or at least, on the surface it insisted on title-casing everything and ignoring any filesystem-defined case.
20170926 17:38:27< DeFender1031> The ONLY concern really is whether the filesystems themselves will allow particular characters.
20170926 17:38:44< DeFender1031> vn971, I vote for both.
20170926 17:39:50< JyrkiVesterinen> shadowm: FAT16 with LFN is case-preserving. It stores file names as UCS-2.
20170926 17:40:01< Necrosporus> JyrkiVesterinen, $ rm #
20170926 17:40:01< Necrosporus> rm: missing operand
20170926 17:40:01< Necrosporus> Try 'rm --help' for more information.
20170926 17:40:12< shadowm> Well, the shell didn't make any attempt to honour that for some inexplicable reason.
20170926 17:40:19< Necrosporus> rm \# works though
20170926 17:40:42< DeFender1031> Necrosporus, shell is irrelevant.
20170926 17:42:49< Soliton> i think disallowing control characters makes sense. the rest is a bit over the top.
20170926 17:44:15< shadowm> I don't think enforcing case on Windows* is over the top if it'll mean less frustrated Linux users giving up on broken add-ons instead of complaining to the author. (* Whether it's feasible or not is not a thing I ever got around to researching.)
20170926 17:44:31< Necrosporus> DeFender1031, JyrkiVesterinen, lets' ask another question. What is the benefit of _allowing_ ascii characters outside of range I mention, namely [A-Za-z0-9+._-] in filenames?
20170926 17:44:48< vn971> https://github.com/wesnoth/wesnoth/issues/2046
20170926 17:45:24< Soliton> sure, i was not commenting on that. if that's feasable that's nice.
20170926 17:45:41< DeFender1031> And *I* don't think that enforcing characters that can actually exist on all systems is over the top either.
20170926 17:46:19< Soliton> do you know a case where we don't already do that?
20170926 17:46:39< vn971> (this is about case-sensitive resource loading) > https://github.com/wesnoth/wesnoth/issues/2046
20170926 17:46:54< DeFender1031> As a single example among many, : doesn't work on FAT32
20170926 17:47:06< DeFender1031> Yet there's no check for it.
20170926 17:47:13< shadowm> (Can somebody remind me how to check if a WML attribute is set and non-empty?)
20170926 17:47:16< Soliton> it is not allowed already.
20170926 17:47:48< DeFender1031> I didn't see that in the code pointed out in the chat above...
20170926 17:48:24< Soliton> i think the actual reason was amiga os.
20170926 17:49:06< Necrosporus> Are addons stored in tar.bz2 archives on server?
20170926 17:49:17< Necrosporus> or archives are regenerated?
20170926 17:49:23< shadowm> They are WML archives.
20170926 17:49:35< shadowm> Just WML documents optionally (?) compressed with gzip.
20170926 17:50:09< Necrosporus> Would it be hard to find how many addons use characters outside of range I mentioned above?
20170926 17:50:23< Necrosporus> Main wesnoth repo has one file with spaces in name
20170926 17:50:33< shadowm> Yes.
20170926 17:50:44< Necrosporus> ./projectfiles/Xcode/Resources/Growl Registration Ticket.growlRegDict
20170926 17:51:27< shadowm> Actually no, it wouldn't be hard since we do keep the unpacked versions around for other purposes, but I think I have better uses for my time.
20170926 17:52:36< Necrosporus> @ is used in context of i18n such as ./doc/man/sr@ijekavian
20170926 17:52:44< Necrosporus> and po file names
20170926 17:53:05< Necrosporus> It's also used in image files, having @2x in front of extension
20170926 17:53:37< JyrkiVesterinen> Necrosporus: the advantage of allowing special characters in filenames is simply giving more freedom to UMC authors.
20170926 17:53:40< Necrosporus> There are 118 @2x files
20170926 17:53:55< JyrkiVesterinen> IMO, if an UMC author wants to name a map, say, 05_Betrayal!.map, let them.
20170926 17:54:44< Necrosporus> Perhaps some of characters could be allowed, which do not have special meaning in WML or Lua
20170926 17:55:08< Necrosporus> = would be undesirable though
20170926 17:55:12< JyrkiVesterinen> But I think having special meanings in WML or Lua is nowhere near big enough reason to outright disallow them in filenames.
20170926 17:55:20< shadowm> ...
20170926 17:55:38< shadowm> = doesn't have a special meaning in a quoted value.
20170926 17:55:48< Necrosporus> OK, but there is also , 
20170926 17:55:51< shadowm> foo="=" is not invalid WML.
20170926 17:56:00< shadowm> foo="," is not invalid WML either.
20170926 17:56:11< Necrosporus> But , delimits list of frames
20170926 17:56:25< shadowm> Not all WML is frame definitions.
20170926 17:56:38< DeFender1031> oh, : is in that code, I missed it the first time. Still, I'm pretty sure there are other characters that FAT32 disallows that aren't in there.
20170926 17:56:51< DeFender1031> I'll check when I get back. Need to run right now.
20170926 17:57:25< Soliton> it seems that so far UMC authors are capable of giving their files names they can handle.
20170926 17:57:46< vn971> JyrkiVesterinen: it's different things, is it? 1. what files can we upload 2. by what "names" can we access files. It seems possible to just check file names themselves, and don't touch all those @2x inside WML files. Or maybe I'm missing something?..
20170926 17:58:34< JyrkiVesterinen> Sure, but as long as the file is accessible easily enough, I find it should be allowed.
20170926 17:58:48< Soliton> @2x is in the filename.
20170926 17:58:59< vn971> Soliton: oh, OK.
20170926 17:59:49< Necrosporus> What do wesnoth resource filenames use @2x suffix?
20170926 18:00:00< Necrosporus> I mean not something else
20170926 18:01:24< shadowm> Why does it matter?
20170926 18:01:42< shadowm> Someone already pointed out there are locale names that have @ in them. It's irrelevant either way.
20170926 18:02:07< JyrkiVesterinen> IIRC, it's a common practice with HiDPI macOS apps. Although I don't know if that where the practice came to Wesnoth.
20170926 18:02:30< JyrkiVesterinen> *that's
20170926 18:02:54< shadowm> https://cdn.discordapp.com/attachments/259976436490829825/362297962753425418/unknown.png
20170926 18:03:03< Necrosporus> All @2x resource files for Wesnoth seems to reside in images or attic directories (not data/images)
20170926 18:03:19< JyrkiVesterinen> shadowm: Looks good enough UX to me. :)
20170926 18:03:33< shadowm> vultraz: Other than fixing that prefix issue (the double forward slashes) I don't intend to do any more client-side work on it.
20170926 18:03:49< shadowm> I'll leave the rest to you. There'll be a comment in the code explaining what I feel is the best way to go about it.
20170926 18:03:58-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20170926 18:04:22< Necrosporus> shadowm, looks OK. Does it also spit out to stdout or stderr of client?
20170926 18:04:23< shadowm> Also I don't think I'll backport this to 1.12.
20170926 18:04:37< Necrosporus> except // prefix
20170926 18:05:05< shadowm> Mhm.
20170926 18:05:14< shadowm> https://pastebin.com/XqCJ36L4
20170926 18:07:26< Necrosporus> It does, that's good
20170926 18:11:27< Necrosporus> Now a question. What if an addon contained a thousand of incorrectly named files
20170926 18:11:41< shadowm> That's why I said I'm leaving the rest to vultraz.
20170926 18:12:08< Necrosporus> I mean, to server side, won't it cause a DoS?
20170926 18:12:22< Necrosporus> I guess probably should not
20170926 18:13:11< shadowm> There are simpler ways to do that.
20170926 18:30:45-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170926 18:31:57< shadowm>     15:29:49 shadowm@hanacore ~/src/wesnoth git:feature/campaignd-2043 % git push -u personal-fork feature/campaignd-2043
20170926 18:32:01< shadowm>     Writing objects:  37% (4023/10872), 5.09 MiB | 81.00 KiB/s
20170926 18:32:10< shadowm> [feeling of profound regret*
20170926 18:39:27< shadowm> Submitted the PR.
20170926 18:41:53< shadowm>     src/tests/test_config.cpp(154): error: in "test_config/test_config_attribute_value": check !cc["x"].empty() has failed
20170926 18:41:56< shadowm>     src/tests/test_config.cpp(158): error: in "test_config/test_config_attribute_value": check !c["x"].empty() has failed
20170926 18:42:03< shadowm> The unit test results look a bit funky on my machine.
20170926 18:42:18< shadowm> I hope that doesn't mean anything. It's not even my code.
20170926 18:43:42< JyrkiVesterinen> WTF? No one has changed config code or config tests recently.
20170926 18:43:52< JyrkiVesterinen> And the tests have continued to pass in Travis, too.
20170926 18:44:28< shadowm> Possibly something specific to my configuration (GCC 7.2.0, Boost 1.62.0).
20170926 18:53:04< Necrosporus> I have just got segfault in wesnoth at very start. libsdl is 2.0.5
20170926 18:53:16< Necrosporus> It didn't repeat on next start
20170926 18:55:02< Necrosporus> version 5e15cdb-Clean, release build
20170926 18:56:46-!- Bhoren [~Bhoren_wh@2a01:e0a:c:2150:f09b:f482:d934:9f68] has joined #wesnoth-dev
20170926 18:58:31< Necrosporus> Any idea what to do with it?
20170926 19:00:08-!- JyrkiVesterinen [~JyrkiVest@85-23-197-3.bb.dnainternet.fi] has quit [Quit: Sleep]
20170926 19:13:52-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20170926 19:20:46-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170926 19:21:20-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170926 19:28:39< vn971> hmm, I just got wesnoth-1.12 crash, at the state of forming room(game) members for a play.
20170926 19:29:14< vultraz> if it's reproducible in master (1.13), do report. else 
20170926 19:29:17< vultraz> don't
20170926 19:29:34< vn971> https://gist.github.com/vn971/4d6ad26170d683479b0bc0dee3142b36
20170926 19:29:50< vn971> vultraz: well there's an assertion that fails, dunno if it's worth investigating.
20170926 19:30:25< vn971> vultraz: it's not reproducable at all because it happened in an MP game, with actions that normally don't lead to bugs. I guess some concurrency thing or smth like that.
20170926 19:31:15< vn971> actually, not even in game. In pre-game room where you configure leaders & sides.
20170926 19:46:21-!- atarocch [~atarocch@93.56.164.28] has quit [Ping timeout: 240 seconds]
20170926 20:05:31-!- Kwandulin [~Kwandulin@pD9FD50F1.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20170926 20:13:46-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170926 20:14:00-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170926 20:54:08-!- travis-ci [~travis-ci@ec2-54-224-236-200.compute-1.amazonaws.com] has joined #wesnoth-dev
20170926 20:54:09< travis-ci> shikadilord/wesnoth#15 (feature/campaignd-2043 - e0d6c28 : Ignacio R. Morelle): The build has errored.
20170926 20:54:09< travis-ci> Build details : https://travis-ci.org/shikadilord/wesnoth/builds/280091849
20170926 20:54:09-!- travis-ci [~travis-ci@ec2-54-224-236-200.compute-1.amazonaws.com] has left #wesnoth-dev []
20170926 21:08:07-!- gfgtdf [~chatzilla@x4e3681d2.dyn.telefonica.de] has quit [Read error: Connection reset by peer]
20170926 21:17:14< shadowm> "The job exceeded the maximum time limit for jobs, and has been terminated." f you too.
20170926 21:36:36-!- travis-ci [~travis-ci@ec2-54-224-236-200.compute-1.amazonaws.com] has joined #wesnoth-dev
20170926 21:36:37< travis-ci> shikadilord/wesnoth#17 (master - 2c20bf7 : Alexander van Gessel): The build has errored.
20170926 21:36:37< travis-ci> Build details : https://travis-ci.org/shikadilord/wesnoth/builds/280099914
20170926 21:36:37-!- travis-ci [~travis-ci@ec2-54-224-236-200.compute-1.amazonaws.com] has left #wesnoth-dev []
20170926 21:47:59-!- boucman [~rosen@wesnoth/developer/boucman] has left #wesnoth-dev []
20170926 22:03:48-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit []
20170926 22:04:06-!- Bhoren [~Bhoren_wh@2a01:e0a:c:2150:f09b:f482:d934:9f68] has quit [Quit: Leaving]
20170926 22:46:04-!- travis-ci [~travis-ci@ec2-54-224-236-200.compute-1.amazonaws.com] has joined #wesnoth-dev
20170926 22:46:05< travis-ci> shikadilord/wesnoth#15 (feature/campaignd-2043 - e0d6c28 : Ignacio R. Morelle): The build passed.
20170926 22:46:05< travis-ci> Build details : https://travis-ci.org/shikadilord/wesnoth/builds/280091849
20170926 22:46:05-!- travis-ci [~travis-ci@ec2-54-224-236-200.compute-1.amazonaws.com] has left #wesnoth-dev []
20170926 23:43:42-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20170926 23:43:48-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
--- Log closed Wed Sep 27 00:00:11 2017