07:39:33 Some feedback here -> https://www.reddit.com/r/Monero/comments/j5qbq1/gui_v01701_oxygen_orion_released/g7wvk09/ 07:47:40 dEBRUYNE: replied, I’m skeptical at that comment 09:31:08 -xmr-pr- endorxmr opened issue #3135: Send screen: some help text does not switch colour when switching from... 09:31:08 -xmr-pr- > https://github.com/monero-project/monero-gui/issues/3135 11:16:09 -xmr-pr- selsta opened pull request #3136: Transfer: fix help text (white theme) 11:16:09 -xmr-pr- > https://github.com/monero-project/monero-gui/pull/3136 14:16:08 -xmr-pr- rating89us opened issue #3137: Feature request: function getApproximateDate() 14:16:08 -xmr-pr- > https://github.com/monero-project/monero-gui/issues/3137 15:11:49 selsta: I'm able to reproduce 15:11:52 "Some problems at connect, message: Socks request rejected or failed" 15:11:58 "Problems at shutdown: Transport endpoint is not connected" 15:12:14 Turning off Socks5 proxy in settings allows wallet to connect to localhost daemon. 15:12:29 I don't have time to debug this any further atm 15:16:57 xiphon: ^ any idea? 15:17:01 tobtoht: thanks 15:24:57 selsta: works as expected 15:25:20 a user tries to connect to a remote node using socks proxy 15:25:28 that's what GUI is doing 15:31:35 ^ They shouldn't use remote node mode to connect to local node with socks proxy enabled 15:32:12 guess that require https://github.com/monero-project/monero-gui/pull/2956 15:32:16 to select custom port on tails? 15:32:18 All we need is just fix local node mode, i.e. to add local rpc bind port configuration and set it to some Tails "allowed" local port 15:33:24 "guess that require https://github.com/monero-project/monero-gui/pull/2956" -> don't like the particular implementation 15:33:52 but yep, basically we need a way to configure local daemon rcp bind port 15:33:56 due to torsocks? 15:34:26 ok 15:36:40 i will have a look a bit later 15:37:19 *maybe* we will fix it in the point release 16:08:51 xiphon: i peeked at the guix package of monero-gui and they are certainly having the same problem i have. see https://git.savannah.gnu.org/cgit/guix.git/tree/gnu/packages/finance.scm#n713. VERSION_IS_RELEASE is being manually patched 16:10:45 could you provide a link to monero source they use? 16:11:05 rnhmjoj: ^ 16:12:28 i think they are just downloading the archive from relase 0.17.0.1 from github 16:12:40 on this line: https://git.savannah.gnu.org/cgit/guix.git/tree/gnu/packages/finance.scm#n672 16:13:44 the sha256 hash is identical to mine, nice 16:14:20 i don't see they fetching monero-gui repo recursively 16:14:39 anyway, i need a steps to reproduce the issue 16:15:09 ^ please open a Github issue with the steps 16:16:48 if you have nix installed it would be pretty simple. otherwise, i guess i could give you a trace of the cmake command invocation and other stuff 16:17:39 "i think they are just downloading the archive from relase 0.17.0.1 from github" 16:17:43 you can't use it 16:18:21 please follow the official building docs and recursively clone monero submodule 16:19:05 at this point it seems to me that you are using outdated monero sources 16:19:08 why not? if you mean because it won't fetch the submodule it doesn't matter because the monero directory is copied from the monero package source, which is itself fetched recursively with git 16:19:57 cause github archives doesn't contain submodules, as you said 16:20:33 alright, we have a CI set up 16:20:53 please have a look at the command we actually run on every commit / pull request 16:21:34 xiphon: the monero sources are from 0.17.0.1 and yes, they contain the submodules all right. in any case the issue is coming from a file within monero-gui source tree, not the monero submodule 16:25:07 it shouldn't use Version.cmake from monero-gui repo, that is clearly a bug 16:25:13 that i fixed a while ago 16:25:29 please let me know if there is something left regarding this 16:25:53 but we already have a PR removing this and all the other unused cmake code from the monero-gui repo 16:26:17 ^ it requires another monero PR to be merged first though 16:26:45 https://github.com/monero-project/monero-gui/pull/3131 16:27:40 uhm, so monero-gui/cmake/verison.cmake should have never been used? 16:27:50 right 16:28:38 rnhmjoj: selsta: sorry, i have to go right now. Will be back in a few hours 16:28:41 * xiphon afk 16:29:09 ok, see you later 16:39:24 rnhmjoj: i will double-check the case, so will remove just Version.cmake from monero-gui repo and see if there is something left that still uses it from monero repo 16:39:27 will let you know 16:39:29 * xiphon afk 16:39:55 xiphon: ok, thank you 20:17:07 rnhmjoj: found it. There is a combination of bugs. Another one: monero's Version.cmake uses wrong version.cpp.in file location (https://github.com/monero-project/monero/pull/6867) 20:18:13 could you please do a test? remove monero-gui/cmake/Version.cmake 20:18:43 and apply the patch from https://github.com/monero-project/monero/pull/6867 to monero-gui/monero/cmake/Version.cmake 20:18:59 xiphon: looks like CI failed 20:19:14 CMake Error: File /home/runner/work/monero/monero/src/src/version.cpp.in does not exist. 20:19:42 on it, will fix in a few seconds 20:21:09 updated 22:45:13 xiphon: that's good news. i'm building monero with the patch rn 22:46:37 rnhmjoj: let me know the results 22:48:35 sure thing. so, the problem is the variable is a reference to the root (which is monero-gui) but it should be monero, right? 23:03:14 xiphon: it seems to have worked. monero-gui is now building without manually patching the IS_RELEASE macro 23:03:48 "the problem is the variable is a reference to the root (which is monero-gui) but it should be monero, right?" -> right 23:04:22 rnhmjoj: thanks for reporting and testing all the stuff 23:06:57 xiphon: no problem, thank you for figuring it out. btw, i still don't understand why no one else noticed this: what exactly is triggering it? 23:11:29 "what exactly is triggering it?" -> didn't dig it further, my best guess is that due to some environment difference monero-gui/cmake directory gets searched first (instead of monero/cmake) from monero-gui/monero/CMakeLists.txt 23:12:20 something like CMAKE_PREFIX_PATH or another similar variables gets set improperty 23:12:54 Make due to different CMake versions (they sometimes change the behaviour between the releases) 23:13:00 What CMake version do you use? 23:13:22 s/Make/Maybe 23:14:22 "i still don't understand why no one else noticed this" -> as i said before, we even have a non-stopping CI running on each commit/PR, works just fine 23:14:39 on 3 operating systems 23:15:14 indeed 23:23:01 xiphon: cmake is 3.18. the explanation it make sense: nix adds a bunch of cmake flags by default and one must be doing it, though it not obvious to me: https://hastebin.com/raw/zedafotani 23:24:38 CMAKE_INSTALL_PREFIX 23:25:13 Looks like this is what breaks the search path 23:26:43 so monero-gui folder will be searched first 23:28:21 that also explain why the guix folk had the same problem. guix is surely setting the prefix in a similar manner 23:31:11 once we merge the patches mentioned earlier, it will work out of the box (by accident though, but still) 23:34:22 yeah, i'm leaving a comment to myself to remove the fix for the next release 23:35:02 xiphon: should we PR these to branch too? 23:35:14 else it will take a while until we can merge it in gui 23:36:56 are there any other fixes for monero core point release? 23:37:20 currently not 23:38:01 not good, point release just for these couple patches doesn't look good 23:38:38 would not do a monero core point release, just PR to branch and once we have more fixes we can do a release sometime in the future 23:40:31 yep, sounds good 23:40:35 I just meant that we won’t branch master to a release branch for at least 6 months or so 23:44:39 yep