[00:21:11] cool will check it out now [01:12:27] was this passing before though right? [01:30:58] seems like the best hash's are both different on each node [01:31:03] but their nbits are the same [01:32:02] not sure what caused this to start failing [01:32:16] must be something from the other branches that i just merged to master? [01:36:09] if i comment out that line, the test passes [01:36:29] so it detects different hashes, then when they are reconnected, they end up on the same hash [01:36:44] so i wonder if the nbits not being equal is a requirement or not [05:46:50] @prole I commented on the github issue [08:31:04] @prole @aguycalled [08:31:09] I've created a new branch [08:31:21] And i'm upgrading QT5 to 5.9.8 [08:31:32] But I've run into and issue with ftgl [08:36:38] what is ftgl exactly? [08:37:02] I was hoping you would know [08:37:10] Wait, let me get an excerpt [08:37:53] OBJCXXLD qt/navcoin-qt CXXLD qt/test/test_navcoin-qt /home/mxaddict/Projects/mxaddict/navcoin-core/depends/x86_64-pc-linux-gnu/share/../lib/libQt5FontDatabaseSupport.a(qfontengine_ft.o): In function `QFontEngineFT::getSfntTableData(unsigned int , unsigned char*, unsigned int*) const': qfontengine_ft.cpp:(.text+0x43b): undefined reference to `FT_Load_Sfnt_Table' [08:37:54] /home/mxaddict/Projects/mxaddict/navcoin-core/depends/x86_64-pc-linux-gnu/share/../lib/libQt5FontDatabaseSupport.a(qfontengine_ft.o): In function `QFontEngineFT::glyphIndex(unsigned int) cons t': qfontengine_ft.cpp:(.text+0x4a6): undefined reference to `FT_Get_Char_Index' qfontengine_ft.cpp:(.text+0x4db): undefined reference to `FT_Set_Charmap' qfontengine_ft.cpp:(.text+0x4e6): undefined reference to `FT_Get_Char_Index' [08:37:54] qfontengine_ft.cpp:(.text+0x4fb): undefined reference to `FT_Set_Charmap' qfontengine_ft.cpp:(.text+0x52d): undefined reference to `FT_Get_Char_Index' qfontengine_ft.cpp:(.text+0x551): undefined reference to `FT_Get_Char_Index' [08:38:05] It's an error related to QT5 [08:38:21] Seems QT 5.9 uses ftgl for fonts [08:38:42] I'm looking at the bitcoin-core codebase now to find where/how they include it [08:39:00] can you point me to your branch? [08:41:34] 1 sec [08:42:06] https://github.com/mxaddict/navcoin-core/tree/patch-5 [08:42:42] ok i see you got also the patches from btc [08:43:22] Yes, I figure those are useful [08:43:46] The errors seems to be related to some font library that might be missing? I'm not sure. [08:44:05] never found this issue [08:44:26] I'm not sure what you mean? [08:48:30] never had to deal with ftgl [08:50:08] Scratch that, I think it's freetype [08:50:19] I sse some other people posted about having issues with this before [08:50:27] I'm still reading the posts, BRB [09:08:51] @aguycalled are you able to build locally with the changes in my branch? Or are you getting the same error? [09:10:11] let me try [09:10:16] need to build your depends [09:10:54] Thanks [09:20:19] it'll take a while [09:31:59] configure fails when using the built depends checking for QTPLATFORM... no configure: error: Package requirements (Qt5PlatformSupport) were not met: No package 'Qt5PlatformSupport' found Consider adjusting the PKG_CONFIG_PATH environment variable if you installed software in a non-standard prefix. Alternatively, you may set the environment variables QTPLATFORM_CFLAGS and QTPLATFORM_LIBS to avoid the need to call pkg-config. [09:32:00] See the pkg-config man page for more details. [09:32:21] I already fixed that [09:32:32] You will need to re-run autogen.sh [09:32:57] ok [09:38:23] seeing lots of this ./qt/paymentrequest.pb.h: In member function ‘payments::Payment* payments::PaymentACK::mutable_payment()’: ./qt/paymentrequest.pb.h:2137:15: error: ‘CreateMaybeMessage’ was not declared in this scope auto* p = CreateMaybeMessage<::payments::Payment>(GetArenaNoVirtual()); ^~~~~~~~~~~~~~~~~~ ./qt/paymentrequest.pb.h:2137:53: error: expected primary-expression before ‘>’ token [09:38:24] auto* p = CreateMaybeMessage<::payments::Payment>(GetArenaNoVirtual()); ^ ./qt/paymentrequest.pb.h: In member function ‘void payments::PaymentACK::set_allocated_payment(payments::Payment*)’: ./qt/paymentrequest.pb.h:2151:47: error: ‘GetOwnedMessage’ is not a member of ‘google::protobuf::internal’ payment = ::google::protobuf::internal::GetOwnedMessage( [09:38:24] ^~~~~~~~~~~~~~~ ./qt/paymentrequest.pb.h: In member function ‘bool payments::PaymentACK::has_memo() const’: ./qt/paymentrequest.pb.h:2164:11: error: ‘_has_bits_’ was not declared in this scope return (_has_bits_[0] & 0x00000001u) != 0; ^~~~~~~~~~ ./qt/paymentrequest.pb.h:2164:11: note: suggested alternative: ‘qHashBits’ return (_has_bits_[0] & 0x00000001u) != 0; ^~~~~~~~~~ [09:38:24] qHashBits [09:38:45] but this is more related to rotobuf than qt i think [09:39:08] Ahh [09:39:15] I got that too before [09:39:27] I needed to run make clean in root of navcoin project [09:39:46] okk [09:40:01] Then, make clean seems to delete some header files [09:40:06] IDK why [09:40:21] But you will have to restore/checkout those missing headers after make clean [09:43:35] it built [09:43:52] No errors? [09:43:54] nope [09:43:59] Must be an issue on my system then [09:44:00] LOL [09:44:04] Darn it [09:44:25] should i trigger a gitian build? [09:44:26] Does it run? 😄 [09:44:31] yes [09:44:51] Nice [09:45:02] I guess so [09:45:10] So we can test it out on the LIVE build system [09:45:26] ok, theres a build still going on from yesterday night [09:45:26] If it works, then I'll create the PR? 😄 [09:45:32] gotta upgrade that tiny vps [09:45:47] No worries [09:45:51] I''m not in a hurry, hahaha [09:46:02] I'm just glad the issue was something in my system [09:46:05] And not actually the code [09:46:18] 😃 [09:47:09] If I can get this running on my machine [09:47:22] I will do a mingw and test on windows as well [10:23:32] @aguycalled can you run find /usr/lib -name libfreetype.so [10:24:02] im using --prefix= [10:24:10] I know, I am as well [10:24:28] But for some reason my build can't find the lib (or so it seems) [10:24:54] /usr/lib/x86_64-linux-gnu/libfreetype.so /usr/lib/x86_64-linux-gnu/vlc/plugins/text_renderer/libfreetype_plugin.so /usr/lib/x86_64-linux-gnu/libfreetype.so.6 /usr/lib/x86_64-linux-gnu/libfreetype.a /usr/lib/x86_64-linux-gnu/libfreetype.so.6.16.1 [10:25:29] find ./depends -name "libfreetype*" [10:25:39] Can you show me the ones in your depends? [10:26:10] ./depends/x86_64-pc-linux-gnu/lib/libfreetype.so.6 ./depends/x86_64-pc-linux-gnu/lib/libfreetype.so ./depends/x86_64-pc-linux-gnu/lib/libfreetype.so.6.13.0 ./depends/x86_64-pc-linux-gnu/lib/libfreetype.la [10:26:49] Mine: /usr/lib/x86_64-linux-gnu/libfreetype.so /usr/lib/x86_64-linux-gnu/libfreetype.so.6.15.0 /usr/lib/x86_64-linux-gnu/libfreetype.a /usr/lib/x86_64-linux-gnu/libfreetype.so.6 /usr/lib/x86_64-linux-gnu/libfreetype.la [10:26:55] For system [10:27:29] And depends: ./depends/x86_64-pc-linux-gnu/lib/libfreetype.so ./depends/x86_64-pc-linux-gnu/lib/libfreetype.so.6 ./depends/x86_64-pc-linux-gnu/lib/libfreetype.so.6.13.0 ./depends/x86_64-pc-linux-gnu/lib/libfreetype.la [10:28:13] this is how i built make clean;./autogen.sh;./configure --prefix=/home/alex/navcoin-core/depends/x86_64-pc-linux-gnu;make -j4 [10:28:35] Yeah, same process as mine [10:28:36] LOL [10:28:41] I'm supper confused [10:28:57] I've even tried deleting the repo [10:29:00] And cloning fresh [10:29:10] I'm confused [10:29:23] yeah weird [10:29:32] lets see what happens with gitian [12:32:01] looks like gitian is not building [12:32:24] oh wait its still going [12:35:59] 💪 [12:40:45] @aguycalled I created a PR, so I can see if it builds on travis-ci [12:43:26] Arg, it does not use QT, LOL [12:45:10] Ohh, some of the builds do 😄 [12:47:53] So on travis, only the darwin/mac build creates QT [12:50:33] gitian is going for it [12:50:56] btw, upgrading to qt 5.9.x implies its not compatible with osx older than 10.10 [12:51:03] that should be added to the release notes in the pr [13:01:08] Ok, will add the note [13:04:32] Looks like the MAC build for QT is working: https://travis-ci.org/NAVCoin/navcoin-core/jobs/527282053 [13:04:47] 🤞 [13:09:30] @aguycalled Looks like the build for MAC with QT 5.9.8 worked https://travis-ci.org/NAVCoin/navcoin-core/jobs/527282053 [13:09:42] I guess the issue really is on my system [13:09:45] Hmmmm 😢 [13:14:20] :/ [13:14:31] sent you the first payment for your first pr merged [13:55:11] gitian did not work [13:55:44] https://cdn.discordapp.com/attachments/416000318149754881/573507920700899338/unknown.png [13:55:50] same as you @mxaddict [14:07:09] Hmmm [14:07:30] What version gcc are running? [14:07:39] And what version is gitlian running? [14:08:05] Maybe it's something to do with the toolchain [14:35:31] 4.8 [14:35:34] https://github.com/NAVCoin/navcoin-core/blob/master/contrib/gitian-descriptors/gitian-linux.yml [14:35:54] i tried once to upgrade it, but it's the newest available in trusty iirc [15:00:50] my laptop has 7.4 [15:02:11] gcc 7.4? [15:02:15] what OS? [15:09:30] @aguycalled [15:10:49] debian [15:13:32] I'm testing with gcc 7.4 now [16:55:05] Seems like it's not the GCC version [16:55:10] 7.4 still fails for me [19:09:26] @aguycalled what version of make do you run on your laptop? [19:11:54] 4.2.1 [19:17:23] I've created a new PR to add the pool op codes to navcore. https://github.com/NAVCoin/navcoin-core/pull/464 [21:04:46] @aguycalled [21:04:58] I just installed debian (docker) [21:05:13] And was able to compile the navcoin-qt with QT 5.9.8 [21:05:23] I'm still not 100% sure what the issue is [21:05:29] Do you have any ideas? [21:06:07] not sure tbh what else could be wrong here [21:06:36] The weird thing is I cloned bitcoin/bitcoin repo [21:06:43] Updated to version 5.9.8 [21:06:51] And it compiled on my main OS [21:06:54] Ubuntu 18.04 [21:07:27] :spazdunno: [21:07:31] weird [21:07:41] used the bitcoin depends? [21:07:46] Yes [21:07:54] I edited bitcoin-core [21:08:05] to use 5.9.8 [21:08:12] Same edits I did to our repo [21:08:16] And it works on bitcoin build [21:08:33] So it must be something different in the way g++ is invoke by our Makefile [21:08:42] VS the bitcoin Makefile [21:09:01] yep thats what i was thinking [21:09:36] i was trying now to find the bitcoin pr which upgraded to 5.9, but i think they upgraded smoother than us [21:12:43] Hmmm, something weird just happend [21:12:52] I built depends in debian docker [21:13:35] Then I did "./configure --prefix=pwd/depends/x86_64-pc-linux-gnu" [21:14:35] And make -j8 on host machine and it builds [21:14:44] I even did a make clean to be sure [21:15:11] I guess the depends built inside depends on out codebase break the lib*.la files [21:23:50] *** Joins: aguycalled (~user@unaffiliated/aguycalled) [21:23:54] *** Server sets mode: +cnt [23:37:41] @here we will have new 4.6.0-rc binaries ready in a couple of hours. I will post the link here, then it would be great if anyone is available to try run it on their system and make sure it works properly. The more people who help to test the RC the quicker we can cut the release and help the exchanges upgrade. [23:39:11] Nice :+1: [23:41:34] Count me in, I'll test it