00:00:50 <Eksell> Stable (0.21) branch on CRAWL.XTAHUA.COM updated to: 0.21.1-0-g3554444
01:10:20 <Gretell> Stable branch on crawl.develz.org updated to: 0.21.1-0-g3554444 (34)
01:24:59 <Gretell> Unstable branch on crawl.develz.org updated to: 0.22-a0-136-g83a10e5 (34)
01:49:16 -!- MadCoyote is now known as FunkyBomb
01:57:31 <Gretell> Windows builds of master branch on crawl.develz.org updated to: 0.22-a0-136-g83a10e5
02:54:46 <Gretell> Monster database of master branch on crawl.develz.org updated to: 0.22-a0-136-g83a10e5
03:12:10 <Rotatell> Stable (0.21) branch on crawl.beRotato.org updated to: 0.21.1-0-g3554444
03:13:49 <Rotatell> Unstable branch on crawl.beRotato.org updated to: 0.22-a0-136-g83a10e5 (34)
07:25:55 <|amethyst> gammafunk:  At least one person has assigned DCSS stuff as homework, search CREDITS for "Oulu University"
07:26:59 <|amethyst> gammafunk:  and google scholar shows several academic publications mentioning "dungeon crawl stone soup",
07:27:18 <gammafunk> that's true, the usability project
07:28:33 <|amethyst> unfortunately I'm not a software engineering person: that seems like the best field for having students work on an existing game with a huge and ancient code base
07:32:18 <|amethyst> I'm pretty sure you don't want my first-year MATLAB students writing Crawl code
07:32:54 <|amethyst> well, some of them would be fine, but not necessarily because of anything they've learned in my class
07:44:51 <wheals> they'd still have more experience than linley did!
08:08:36 <|amethyst> wheals:  that's not fair, Linley read a few chapters of a C++ manual :)
13:05:12 <espais> !tell Napkin i was poking around with some single sign-on approaches and was curious about your thoughts on them
13:05:13 <Sequell> espais: OK, I'll let napkin know.
13:05:40 <espais> Napkin: basically, the two options i see are to implement an openID implementation (people click a button and login with google/twitter/facebook/steam/etc)
13:06:31 <espais> Napkin: or, a master server (probably hosted by you i'd guess as part of the mantis server) generates a token that could be sent to each server
13:09:20 <espais> Napkin: alternatively we could host our own openID provider if we wanted to avoid 3rd party providers, but with openID becoming pretty common for logging i i thought i'd toss it out there
13:11:58 <espais> i think what i'll do is dink around with it a bit on my private instance and see what i come up with ... my biggest concern would be how something like SSO would impact existing logins
13:12:25 <geekosaur> that has been the sticking point, not all accounts match across servers
13:12:42 <geekosaur> that is, some accounts that exist on multiple servers do not represent the same user
13:15:18 <espais> yea i think there would have to be some sort of transition process for that
13:16:04 <espais> ugh i didn't even think of that, that would be a messy thing
13:17:38 <espais> so if we went sso, do you think users would accept changing their usernames to be unique across all domains?
13:17:46 <espais> or is this just a pipe dream
13:21:04 <wheals> i think people would be happy with it after the fact, the main problem is the migration
13:21:42 <geekosaur> yeh. which is a people problem more than a tech problem so everyone's /o\ about it
13:22:35 <espais> hmm, i wonder if aliases would be acceptable then
13:23:12 <espais> so i'd login as espais as a primary account name but could specify my alt per server? naturally on cbro i'm ultraviolent4
13:23:18 <espais> >_>
13:23:34 <geekosaur> I'd consider per-server maps and generated account ids
13:25:04 <espais> yea
13:25:14 <espais> this seems messy at a people level
13:25:15 <espais> like you said
13:25:20 <espais> i can see why it hasn't happened yet
13:30:05 <espais> maybe i'll stick to dinking around with getting into map making for now
14:52:16 <espais> interesting
14:52:24 <espais> so i just pulled a fresh copy of crawl from git (cygwin)
14:52:30 <espais> ran git submodule update --init and all that
14:52:36 <espais> pulled in all packages
14:52:40 <espais> got a bunch of compilation errors
14:52:47 <espais> cio.cc: In member function ‘int line_reader::read_line_core(bool)’: cio.cc:471:26: error: ‘wcwidth’ was not declared in this scope          cpos += wcwidth(c);                           ^ cio.cc: In member function ‘void line_reader::backspace()’: cio.cc:632:39: error: ‘wcwidth’ was not declared in this scope      const int glyph_width = wcwidth(ch);                                        ^ cio.cc: In member funct
14:53:04 <espais> bit cleaner, sorry
14:53:05 <espais> http://paste.debian.net/1008580/
14:56:20 <espais> (packages were pulled in via cygwin's setup -q -P (the laundry list of linux-required packages from apt-get)
14:57:06 <espais> trying with no_pkgconfig next
15:00:50 <geekosaur> hm
15:00:59 <geekosaur> I think this might actually be a crawl bug
15:01:12 <geekosaur> on windows native, crawl provides its own wcwidth()
15:01:18 <geekosaur> on unix, it relies on the OS one
15:01:36 <geekosaur> but the manpage says you need to #include <wchar.h> to get it
15:02:00 <geekosaur> libunix.cc pulls that in (via C++ <cwchar>), nothing else seems to
15:03:31 <wheals> i guess it gets pulled in transitively on non-cgywin unixes, then
15:03:56 <wheals> i wonder when it broke since i don't remember any recent changes to cio.cc and people have compiled on cygwin in the past
15:04:50 <espais> yea first time compiling for me on windows since switching from ubuntu
15:05:52 <espais> interestingly, with NO_PKGCONFIG
15:05:53 <espais> x86_64-pc-cygwin-g++: error: ../contrib/install/x86_64-pc-cygwin/lib/libpng.a: No such file or directory x86_64-pc-cygwin-g++: error: ../contrib/install/x86_64-pc-cygwin/lib/libz.a: No such file or directory
15:06:46 <geekosaur> cygwin would expect it to come from the OS and not build the contribs, then the build doesn't see it from pkgconfig and assumes the contribs will provide it
15:07:08 <geekosaur> another of the "lovely" ways the build system trips over its own feet
15:09:07 <espais> searching for libpng.a now, i thought i'd installed libpng-devel
15:50:44 <espais> i wanted to give Dith some love
15:50:44 <espais> probably a silly question, but i was just playing with adding a simple float map
15:50:45 <espais> do i only need to update float.des or is there something else that needs to be done (eg recompile...)
15:50:45 <espais> http://paste.debian.net/1008588/
15:51:05 <espais> going into wizmode it won't let me add in espais_dith_palace as the name cannot be found
15:51:19 <espais> (the map in the paste above is just seeing if i can spawn something)
15:53:44 <espais> just realized DESC was mispelled but that didn't break it
15:54:39 <espais> i'm an idiot
15:54:40 <geekosaur> you have to tebuild the des cache
15:54:49 <espais> i'm editing my git files, not my windows build
15:54:56 <geekosaur> oh, whoops, yes, have to install it :)
15:55:15 <espais> how does one rebuild the cache? didn't see a note in the mapmaking guide
15:56:27 <geekosaur> normally it's done when you start a new game, but there have been cases when it didn't happen
15:56:36 <espais> ok
15:56:51 <espais> that did it
15:57:04 <espais> one thing i don't understand from the glyph guide if you don't mind
15:57:16 <espais> so B adds an altar, how does one specify a particular altar (ie dith)
16:02:28 <geekosaur> oh, --builddb was the option to force rebuiulding the des cache
16:03:32 <geekosaur> KFEAT: B = altar_dithmenos
16:05:19 <espais> thx
16:05:24 <espais> this...is addicting
16:13:43 <espais> so if i wanted to do something like place a randart quickblade
16:14:28 <espais> i'd place an item and then do something like a KITEM call?
16:15:10 <geekosaur> usually just ITEM
16:15:39 <geekosaur> portals/wizlab.des:1534: [wizlab_cloud] ITEM:       randart ring of flight / good_item boots ego:flying / good_item boots ego:running / good_item quick blade / good_item ring of evasion
16:16:37 * geekosaur needs to figure out the remaining bugs in vaultgrep. it's already on github, at least, but sometimes fails to find things it ought to, plus it's a bit inflexible
16:17:05 <espais> !beer geekosaur
16:17:05 * Sequell slides a mug of root beer down the bar to geekosaur, courtesy of espais.
16:55:44 -!- amalloy_ is now known as amalloy
17:05:03 <espais> ok silly question
17:05:24 <espais> i saw the fog machine code in the wizlab and wanted to include that, but whenever i include the following crawl crashes on loading my vault
17:05:33 <espais> MARKER:     ! = lua:fog_machine { \                 pow_max = 10, delay_min = 10, delay_max = 40, \                 size = 2, size_buildup_amnt = 4, \                 size_buildup_time = 25, cloud_type = "white fluffiness" \             }
17:05:50 <espais> is fog_machine something i'd need to include?  i didn't notice a ref in the wizlab code
17:08:51 <geekosaur> do you have a ! symbol in the map?
17:08:57 <espais> yes
17:09:14 <advil> I think all those wcwidth calls are ones introduced in ea5fd92a9cfc1807a6baec22cf856402dcecd618
17:09:17 <advil> %git ea5fd92a9cfc1807a6baec22cf856402dcecd618
17:09:17 <Cheibriados> 07advil02 * 0.21-a0-301-gea5fd92: Add UI for skill target editing for console/local tiles 10(4 months ago, 16 files, 836+ 143-) 13https://github.com/crawl/crawl/commit/ea5fd92a9cfc
17:09:31 <advil> as to why they aren't compiling in cygwin specifically, no idea (since msys2 compiles fine)
17:09:43 <espais> width
17:09:44 <espais> that's it
17:09:56 <espais> i was putting it in a space that it couldn't fit
17:10:31 <advil> is it crashing on an assert?
17:11:55 <wheals> msys2 and cywgin are pretty different
17:11:55 <wheals> so it's not surprising to me that only one would work
17:12:35 <geekosaur> I would suspect, per the manpage, that just dropping a "#include <cwchar>" in the right place would work
17:12:48 <espais> actually hmm, the map crashing thing is too many !'s
17:12:49 <geekosaur> might look to see when/why the one in libunix.cc was added
17:33:59 <espais> !dith
17:43:42 <espais> ok last map q for the day
17:43:49 <espais> i added TAGS:   no_monster_gen no_item_gen no_pool_fixup
17:43:58 <espais> yet i see items and monsters still being generated in the vault
18:02:55 -!- amalloy is now known as amalloy_
18:08:14 <espais> geekosaur: are you saying simply commenting out the cwchar include would work, or should it be replaced?
18:08:23 <espais> because i commented it out and it still horribly broke during the compile
18:08:31 <geekosaur> espais, I am saying it is *needed* elsewhere
18:08:39 <geekosaur> not that it is the cause
18:08:48 <espais> gotcha
18:09:01 <espais> oh dropping it in
18:09:05 <espais> not dropping it out
18:09:06 <espais> :/
18:09:06 <geekosaur> wcwidth() is only guaranteed to be available in C++ if <cwchar> has been included
18:09:09 <|amethyst> probably unicode.h makes the most sense
18:09:16 <espais> <-- reading comprehension
18:09:32 <|amethyst> since that's where the non-unix declaration of the crawl wcwidth implementation is
18:10:11 <geekosaur> also possibly there is a #define needed but I have no idea how it would interact with cygwin (_XOPEN_SOURCE)
18:11:04 <|amethyst> hm, I forget, is it safe to #define that before #including cwchar then undefine it afterwards?
18:12:54 <geekosaur> I have no idea and the answer might differ on different platforms anyway
18:13:00 <geekosaur> "portability"
18:13:07 <geekosaur> !flip portability
18:13:07 <Sequell> (╯°□°)╯︵ʎʇᴉlᴉqɐʇɹ°d
18:14:27 <|amethyst> "In the compilation of an application that #defines a feature test macro specified by POSIX.1-2008, no header defined by POSIX.1-2008 shall be included prior to the definition of the feature test macro."
18:14:32 <|amethyst> ugh
18:14:50 <|amethyst> so _XOPEN_SOURCE would need to be at the top of AppHdr.h
18:14:54 <|amethyst> or near the top
18:15:29 <geekosaur> (especially since, if you need wcwidth(), chances are you stopped being portable anyway. it depends on the output device, nit just the codepoint and locale!)
18:21:09 <|amethyst> geekosaur:  yeah, trees and arches look funny for some people because they're halfwidth according to glibc and its usual en_US.UTF-8 locale, but many fonts make them fullwidth
18:21:25 <geekosaur> yep. see that one all the time here
18:21:38 <geekosaur> actually arches are fine in my font but trees are doublewidth
18:21:51 <geekosaur> or fullwidth as unicode puts it
18:21:56 <geekosaur> 2 terminal cells
18:41:03 <Lightli> I wish there was a graphical way to edit vaults
18:41:26 <Lightli> like vaultedit before the site hosting it murdered it due to a policy change
18:42:45 -!- DarkBattle is now known as AngelaSmythe
19:37:34 <advil> re the wcwidth thing, it's also slightly puzzling why only cio.cc has an error
19:37:51 <advil> that library must still be getting pulled in somehow in a few other places in cygwin
19:39:11 <advil> e.g. initfile.cc
19:39:21 <advil> maybe <string> would load it? or <cstdio>? I can't find anything that indicates that would be standard though
19:43:50 <geekosaur> <ios>, looks like (common defs for istream and ostream)
19:44:00 <geekosaur> possibly an unintended leak
19:44:08 -!- ProzacElf_ is now known as ProzacElf
19:44:40 <geekosaur> {i,o}stream -> ios -> bits/char_traits.h -> cwchar
19:45:07 <geekosaur> and this stands a decent chance of depending on the g++ version
19:46:20 <geekosaur> or more precisel libstdc++ version
21:15:33 <Rotatell> dylex (L20 GnWn) ERROR in 'libutil.cc' at line 367: screen write out of bounds: (52,15) into (42,51) (WizLab)
22:29:19 -!- ProzacElf_ is now known as ProzacElf