[00:16] *** balrog has joined #jsmess [04:24] *** davidar has joined #jsmess [07:07] *** gamingrob has quit IRC (Read error: Connection reset by peer) [07:09] *** gamingrob has joined #jsmess [08:38] *** Ctrl-S___ has joined #jsmess [08:58] *** balrog has quit IRC (Read error: Operation timed out) [08:59] *** balrog has joined #jsmess [09:13] *** datajerk has quit IRC (Ping timeout: 603 seconds) [09:13] *** datajerk has joined #jsmess [10:13] *** godane has quit IRC (Leaving.) [11:06] *** godane has joined #jsmess [14:52] *** db48x has joined #jsmess [14:52] DFJustin: ping? [15:15] pong [15:19] DFJustin: I'm getting some undefined symbols when compiling f-32 [15:19] [2018-07-02 14:55:21] Building mamef-32.js.gz failed with errors [15:19] [2018-07-02 14:58:43] warning: unresolved symbol: _ZN18diagnostics_module12get_instanceEv [15:19] [2018-07-02 14:58:43] warning: unresolved symbol: DEBUG_NONE [15:19] [2018-07-02 14:58:43] warning: unresolved symbol: MIDI_NONE [15:19] [2018-07-02 14:58:43] warning: unresolved symbol: NETDEV_NONE [15:19] [2018-07-02 14:58:43] warning: unresolved symbol: OUTPUT_NONE [15:20] DFJustin: can you recommend a fix? [15:21] first thing is to see if it happens with a native compile [15:36] successfully did a tiny native build of f-32 here with latest git [15:36] so either it's been fixed already or there's some emscripten fuckery [15:50] no unresolved symbols with emscripten and latest git, trying 0.199 [15:51] yea, it works here too [15:52] I'm on the mame0198 branch [15:52] tag [16:04] no unresolved symbols (other than the gl stuff) with mame0199 tag [17:02] I get the same unresolved symbols from mame0199 [17:38] those symbols are defined in os-specific files so somehow your build is not going down the correct os path [18:00] *** azakai has quit IRC (Read error: Operation timed out) [18:38] ah [18:39] hrm [18:39] DFJustin: this doesn't look os-specific [18:39] db48x@glowing-computing-machine:/src/mame/src$ grep -r MIDI_NONE [18:39] osd/modules/lib/osdobj_common.cpp: REGISTER_MODULE(m_mod_man, MIDI_NONE); [18:39] osd/modules/midi/none.cpp:MODULE_DEFINITION(MIDI_NONE, none_module) [18:39] *** pfalleno1 has quit IRC (Read error: Connection reset by peer) [18:40] *** pfallenop has joined #jsmess [18:40] *** pfallenop has quit IRC (Read error: Connection reset by peer) [18:43] do you have these files: [18:43] $ ls build/asmjs/obj/Release/osd_sdl/src/osd/modules/midi/ [18:43] none.d none.o portmidi.d portmidi.o [18:45] yep [18:46] *** pfallenop has joined #jsmess [18:46] if you delete them, does the build recreate them [18:46] *** pfallenop has quit IRC (Read error: Connection reset by peer) [18:49] checking [18:51] *** pfallenop has joined #jsmess [18:51] *** pfallenop has quit IRC (Read error: Connection reset by peer) [18:54] *** pfallenop has joined #jsmess [18:54] *** pfallenop has quit IRC (Read error: Connection reset by peer) [18:59] *** pfallenop has joined #jsmess [18:59] *** pfallenop has quit IRC (Remote host closed the connection) [19:18] odd [19:18] it does rebuild it, and I no longer get a complaint about MIDI_NONE [19:19] maybe if I just clean the build and try again :P [19:55] that fixed it [20:06] there is a known issue with the windows build where if you interrupt the make process, it leaves behind 0-byte .o files which make counts as "done" and the linker happily accepts until it can't find the symbols at the end [20:06] maybe something similar happened [21:09] *** azakai has joined #jsmess [21:11] *** godane has quit IRC (Read error: Operation timed out) [21:37] hmm [21:38] that's not an exact match, but I guess it's possible [21:41] disk full can also cause that I think [21:42] that has happened a few times