2020-05-11 GnuCash IRC logs
00:06:52 *** Mechtilde has joined #gnucash
00:09:22 *** jervin has joined #gnucash
00:10:33 *** jervin has quit IRC
00:55:52 *** angel has quit IRC
01:06:38 *** mohave has joined #gnucash
01:09:38 *** mohave has quit IRC
01:18:54 *** sbluhm has joined #gnucash
01:18:54 *** ChanServ sets mode: +v sbluhm
01:22:49 *** linas has quit IRC
01:23:02 *** linas has joined #gnucash
01:39:09 *** sbluhm has quit IRC
01:40:11 *** fell_laptop has joined #gnucash
01:40:11 *** ChanServ sets mode: +o fell_laptop
01:42:23 *** fell has quit IRC
01:42:35 *** Gerd has joined #gnucash
01:44:05 *** Mechtilde has quit IRC
01:56:39 *** sbluhm has joined #gnucash
01:56:40 *** ChanServ sets mode: +v sbluhm
02:17:19 *** linas has quit IRC
02:18:13 *** linas has joined #gnucash
02:20:20 *** aminbegood has joined #gnucash
02:36:23 *** bertbob has quit IRC
02:44:32 *** bertbob has joined #gnucash
02:44:33 *** ChanServ sets mode: +v bertbob
03:04:11 *** gjanssens has joined #gnucash
03:04:11 *** ChanServ sets mode: +o gjanssens
03:10:17 *** aminbegood has quit IRC
03:14:01 *** Gerd has quit IRC
03:14:10 *** Gerd has joined #gnucash
03:20:01 *** mohave has joined #gnucash
03:21:38 *** jervin has joined #gnucash
03:22:57 *** jervin has quit IRC
03:23:01 *** mohave has quit IRC
03:37:52 *** fell_laptop has quit IRC
03:38:05 *** fell_laptop has joined #gnucash
03:38:06 *** ChanServ sets mode: +o fell_laptop
03:39:59 *** Gerd has quit IRC
03:40:55 *** suukim has joined #gnucash
03:53:48 *** Gerd has joined #gnucash
04:36:29 *** aminbegood has joined #gnucash
04:52:06 *** gjanssens has quit IRC
05:30:14 *** Mechtilde has joined #gnucash
05:54:55 *** fell_laptop is now known as fell
06:02:03 *** User_ has joined #gnucash
06:17:33 *** Aussie_matt has quit IRC
07:06:19 *** mohave has joined #gnucash
07:09:20 *** mohave has quit IRC
07:45:41 *** gour has joined #gnucash
07:45:41 *** ChanServ sets mode: +v gour
07:51:22 *** Gerd has quit IRC
07:55:25 *** gjanssens has joined #gnucash
07:55:25 *** ChanServ sets mode: +o gjanssens
08:20:59 *** Gerd has joined #gnucash
08:21:16 *** aminbegood has quit IRC
08:21:50 *** aminbegood has joined #gnucash
08:40:19 *** chris has joined #gnucash
08:40:19 *** ChanServ sets mode: +v chris
08:50:14 *** Gerd has quit IRC
09:00:09 *** lmat has joined #gnucash
09:08:23 *** crossan007 has quit IRC
09:08:25 *** crossan007 has joined #gnucash
09:13:08 <chris> I don't think there is *ever* a use case for showing monetary amounts using fractions isn't it? so, if a monetary amount is an exact fraction then we should always convert to decimals?
09:13:08 <gncbot> chris: Sent 16 hours and 56 minutes ago: <fell> Isn't 'Bug 569749 - Company name should populate corresponding fields in reports' already fixed?
09:22:54 *** sbluhm has quit IRC
09:40:41 <warlord> chris, IMHO, correct; I see no reason to ever display a fraction.
09:41:15 <chris> warlord: then you have blessed https://github.com/Gnucash/gnucash/pull/720 :)
09:42:24 <warlord> In principle, yes. I haven't looked at the code.
09:46:20 <chris> maybe bugzilla needs a shadow-ban option
09:49:12 *** gour has left #gnucash
09:49:54 <warlord> shadow-ban option?
09:51:46 *** mohave has joined #gnucash
09:53:56 <chris> for the mosquito in 113772
10:01:39 <warlord> I don't agree with the bug, but we do need "tags"
10:01:46 <warlord> (aka classes)
10:02:35 *** Gerd has joined #gnucash
10:06:55 *** Hamaryns has joined #gnucash
10:06:55 *** ChanServ sets mode: +v Hamaryns
10:12:26 *** storyjesse has joined #gnucash
10:18:41 <warlord> chris, and just received private email about that bug.
10:28:47 *** storyjesse1 has joined #gnucash
10:28:59 *** storyjesse has quit IRC
10:29:00 *** storyjesse1 is now known as storyjesse
10:30:37 <chris> warlord: you could forward to core
10:31:39 *** storyjesse1 has joined #gnucash
10:31:59 *** storyjesse has quit IRC
10:32:00 *** storyjesse1 is now known as storyjesse
10:35:18 <warlord> chris, good idea.
10:38:05 *** sbluhm has joined #gnucash
10:38:05 *** ChanServ sets mode: +v sbluhm
10:40:31 *** Mechtilde has quit IRC
10:56:47 *** aminbegood has quit IRC
10:56:50 <chris> jralls: master is still defaulting to dumping too much onto tracefile
11:00:42 *** Gerd has quit IRC
11:01:52 <warlord> Too much is better than too little ;)
11:03:54 *** waeking has quit IRC
11:06:13 *** waeking has joined #gnucash
11:06:13 *** ChanServ sets mode: +v waeking
11:09:29 *** phoenix has joined #gnucash
11:12:35 *** sbluhm has quit IRC
11:15:23 *** Gerd has joined #gnucash
11:20:13 *** jervin has joined #gnucash
11:25:55 <warlord> unfortunately editing comments is a BZ6 feature (which hasn't been released), so that's not a direct option.
11:29:53 *** sbluhm has joined #gnucash
11:31:52 *** omnireq has quit IRC
11:32:03 *** omnireq has joined #gnucash
11:32:03 *** ChanServ sets mode: +v omnireq
11:37:47 *** jervin1 has joined #gnucash
11:38:13 *** jervin has quit IRC
11:38:13 *** jervin1 is now known as jervin
11:38:35 <chris> ideally it would be hellbanning
11:48:36 *** angel has joined #gnucash
11:49:02 *** storyjesse has quit IRC
11:59:08 *** sbluhm has quit IRC
12:04:11 *** angel has quit IRC
12:05:33 *** storyjesse has joined #gnucash
12:08:09 *** Hamaryns has quit IRC
12:09:02 *** bertbob has quit IRC
12:23:57 *** Gerd has quit IRC
12:28:41 *** bertbob has joined #gnucash
12:28:42 *** ChanServ sets mode: +v bertbob
12:33:04 <jralls> warlord, chris, see my response to warlord on core.
12:34:54 *** ArtGravity has joined #gnucash
12:34:54 *** ChanServ sets mode: +v ArtGravity
12:35:33 <jralls> chris, Amounts and values should be displayed as decimals (good thing we don't have £/s/d anymore!). Prices can be rounded or exact depending on a preference. There's already C code for that, the scheme should just call it.
12:36:05 <mohave> jralls, regarding your comments from yesterday, we were indeed on different pages. I was able to get libgnc-engine.dylib built by manually running the compile command and adding the following to it: -lboost_chrono -lboost_date_time -lboost_filesystem -lboost_locale -lboost_regex -lboost_system -lboost_thread
12:36:36 <jralls> chris, I noticed that the next function after the one you modified in PR 720 uses double-to-numeric. Where is the double coming from?
12:41:09 *** guak has joined #gnucash
12:44:02 *** jervin has quit IRC
12:44:18 <jralls> mohave, that's weird. The link command ninja creates lists the boost libraries with full paths right after lib/libgnc-core-utils.dylib and before the first -L. Does it not do that for the xcodebuild link command?
12:45:51 <jralls> mohave, combined with mta's observation about gnc-version-info.h the other day on gnucash-devel I'm leaning towards thinking that maybe Cmake's Xcode generator has problems.
12:46:33 <mohave> Actually I switched to ninja, just to reduce the number of variables.., so this is with ninja.
12:49:45 <jralls> Oh? If you run `ninja -nv gnc-engine` are the boost libraries in the link command?
12:50:53 <chris> jralls: in PR720 the double-to-numeric is called for numbers being sent to reports -- e.g. (gnc:html-document-add-object! document (list (gnc:make-gnc-monetary USD 2.50) 3)) would lead to "$2.50 3.00"
12:51:30 <chris> usually raw numbers shouldn't be rendered; the double-to-numeric becomes the default renderer
12:52:13 <chris> so it's not really
12:52:38 <chris> "double"->"gnc_numeric"... it's really (number? x) --> render as number, with 2 decimal spaces
12:52:38 <jralls> chris: Usually doubles shouldn't happen. I've explained why before.
12:53:35 <mohave> jralls, they are, but listed like: /Users/ctest/gnucash-unstable/lib/libboost_date_time.dylib instead of -lboost_date_time
12:54:50 <jralls> Right. That's faster because the linker doesn't have to search for the library.
12:54:53 <chris> jralls: I'm 99% sure converting the double-to-gnc-numeric to (gnc-numeric-convert NUM 100 GNC-HOW-RND-ROUND) would give the exact same result
12:56:24 <mohave> jralls, okay, so what are your thoughts on why what I did made things work?
12:57:13 <jralls> chris: That's only true for cases where the double is able to exactly represent the number. So 3.00 is fine, 3.10 not so much... though if you then round the double so that only three digits are displayed it will seem to be OK.
12:57:49 <jralls> chris: More than 6 digits and things go south. Scheme has exact decimals and rationals. Use them.
12:58:29 <chris> jralls: not sure I follow. the default-number-renderer isn't used very much. maybe in advanced-portfolio.scm which is still opaque to me.
12:58:31 <jralls> mohave: A more useful question is why is it not working without, since it does work for me.
12:58:46 <chris> jralls: "Prices can be rounded or exact depending on a preference" --> then #720 is incorrect
12:59:24 <jralls> chris: If it's only in APR then it's probably OK. That's under mta's care and he understands doubles.
13:00:17 <jralls> chris: I wondered about that but haven't yet had a chance to dig into it. Like I said, there's already C code that handles it for the UI, you just need to call that for reports.
13:00:48 <jralls> mohave, do any of the libboost paths begin with @rpath?
13:01:20 <chris> jralls: ok. then 797743 is a wontfix. unless I hunt every possible "price"-type monetary display and use the C code instead.
13:02:26 <mohave> jralls, Why is it not working for me, since it does for you? :-)
13:02:36 <jralls> chris, just because you don't want to fix it now doesn't make it a wont-fix.
13:03:00 <mohave> no, no paths begin with @rpath.
13:03:57 <chris> gtg now. if you remember what is the C price renderer let me know...
13:04:37 <jralls> Were the missing symbols directory_iterator and friends?
13:05:33 <mohave> They were like:
13:05:38 <mohave> Undefined symbols for architecture x86_64:
13:05:39 <mohave> "boost::regex_error::regex_error(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, boost::regex_constants::error_type, long)", referenced from:
13:05:39 <mohave> boost::re_detail_107300::basic_regex_parser<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::fail(boost::regex_constants::error_type, long, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, long) in gnc-datetime.cpp.o
13:05:40 <mohave> boost::re_detail_107300::basic_regex_creator<char, boost::regex_traits<char, boost::cpp_regex_traits<char> > >::fixup_recursions(boost::re_detail_107300::re_syntax_base*) in gnc-datetime.cpp.o
13:05:47 *** Hamaryns has joined #gnucash
13:05:47 *** ChanServ sets mode: +v Hamaryns
13:08:09 <jralls> OK, interesting. And /Users/ctest/gnucash-unstable/lib/libboost_regex.dylib is in the ninja output?
13:09:16 <mohave> of: ninja -nv gnc-engine |grep boost, yes
13:13:14 <mohave> Here is the output: https://pastebin.com/ss1U5AXZ
13:16:05 <jralls> What does `otool -L bin/libgnc-engine.dylib | grep libboost_regex` return?
13:17:33 <mohave> I think you meant lib
13:17:38 <mohave> build % otool -L lib/libgnc-engine.dylib | grep libboost_regex
13:17:38 <mohave> /Users/ctest/gnucash-unstable/lib/libboost_regex.dylib (compatibility version 0.0.0, current version 0.0.0)
13:18:22 *** sbluhm has joined #gnucash
13:24:21 <jralls> Wait a minute. Something's really messed up there: you told it to build gnc-engine but that link line is for gnc-core-utils.
13:26:28 <mohave> I see what you are talking about
13:27:19 <jralls> Ah, it's because you pasted step 4 instead of step 5.
13:27:55 <jralls> Step 4 presumably worked since your linker errors were in gnc-date-time and that's in engine.
13:28:46 <mohave> Would you like to see step 5?
13:28:53 <jralls> Yes,
13:29:55 <mohave> https://pastebin.com/zajCUuA0
13:30:31 <jralls> No libboost paths.
13:31:38 <mohave> That's what I saw earlier, which prompted me to add them to the commandline
13:36:35 *** jervin has joined #gnucash
13:37:58 <mohave> Interesting ninja output, as a novice, I would have expected the output of step 5 of 5 to appear after 4 of 5... but alas... guess that's what happens when things run in parallel.
13:38:05 <jralls> What cmake version do you have?
13:38:33 <mohave> build % cmake --version
13:38:34 <mohave> cmake version 3.14.5
13:38:34 <mohave> CMake suite maintained and supported by Kitware (kitware.com/cmake).
13:38:35 <mohave> build %
13:39:01 <jralls> OK.
13:39:20 *** suukim has quit IRC
13:40:18 <jralls> Yes, 4/5 should go before 5/5. Another weirdness.
13:42:36 <jralls> In libgnucash/engine/CMakeLists.txt can you change lines 220 and 221, replacing Boost_FOO_LIBRARIES with Boost_FOO_LIBRARY and rerun ninja -vn gnc-engine?
13:43:06 *** sbluhm has quit IRC
13:45:52 *** jervin has joined #gnucash
13:52:07 *** storyjesse has quit IRC
13:55:19 <mohave> Urg, what am I doing wrong?
13:55:25 <mohave> build % ninja -nv gnc-engine
13:55:25 <mohave> [0/1] /Users/ctest/gnucash-unstable/bin/cmake -S/Users/ctest/Source/gnucash/gnucash -B/Users/ctest/Source/gnucash/build
13:55:25 <mohave> build %
13:58:28 <jralls> Sorry, my mistake. "n" means "tell me what you'd do but don't do it", so it's telling you that it would run cmake. Which is good, but you need it to actually run cmake, so you need to leave off the 'n'.
14:00:42 <mohave> A successful build: -rwxr-xr-x 1 ctest staff 4525236 May 11 10:59 libgnc-engine.dylib
14:01:46 <jralls> Super.
14:05:56 *** Gerd has joined #gnucash
14:09:23 *** Hamaryns has quit IRC
14:10:48 *** frakturfreak has joined #gnucash
14:10:49 *** ChanServ sets mode: +v frakturfreak
14:10:57 *** mohave has quit IRC
14:31:11 *** mohave has joined #gnucash
14:34:11 *** mohave has quit IRC
14:40:08 *** chris has quit IRC
14:46:31 *** chris has joined #gnucash
14:46:31 *** ChanServ sets mode: +v chris
14:51:51 *** Hamaryns has joined #gnucash
14:51:51 *** ChanServ sets mode: +v Hamaryns
15:00:44 *** gjanssens has quit IRC
15:05:27 *** gjanssens has joined #gnucash
15:05:27 *** ChanServ sets mode: +o gjanssens
15:09:23 *** Gerd has quit IRC
15:14:28 *** gjanssens has quit IRC
15:18:43 *** Gerd has joined #gnucash
15:22:08 *** mohave has joined #gnucash
15:33:23 *** sbluhm has joined #gnucash
15:33:23 *** ChanServ sets mode: +v sbluhm
15:37:28 <mohave> jralls: continuing with build, encountered this error:
15:37:30 <mohave> https://pastebin.com/L5RDC40y
15:37:59 *** aminbegood has joined #gnucash
15:38:42 <mohave> Prior to building, I did a git pull, hoping to update my tree with any recent changes
15:39:38 <jralls> That's the one I've been having trouble with off-and-on. I keep thinking I've gotten it fixed only for it to pop up again.
15:42:24 <jralls> No, wait, I did figure that out yesterday, I just haven't figured out how to fix it. The problem is that clang-darwin links libboost dependencies with @rpath and libltdl (the gnu-libtool dynamic loader) doesn't know what to do with it.
15:43:51 <jralls> I worked around it yesterday with `install_name_tool -change @rpath/libboost_threads.dylib `pwd`/libboost_threads.dylib libboost_locale.dylib` after cding to $PREFIX/lib.
15:44:08 <jralls> But that's not really a fix.
15:45:15 <jralls> The offending code is at https://github.com/boostorg/build/blob/develop/src/tools/clang-darwin.jam#L188.
15:50:23 <jralls> It might not be ltdl's fault, either. `otool -l libboost_locale.dylib | grep RPATH` doesn't return anything, meaning that they didn't set an rpath, they're just messing up the install name.
15:51:34 *** Hamaryns has quit IRC
15:57:02 <mohave> did you mean 'install_name_tool -change @rpath/libboost_threads.dylib `pwd`/libboost_threads.dylib' and this command for each boost lib?
15:59:21 <jralls> Each of the ones that shows up for `otool -L libboost_locale.dylib | grep libboost`, yes.
16:00:32 <jralls> And note that it's install_name_tool -change <old-path> <new-path> <library-whose-paths-need-changing>
16:07:45 *** aminbegood has quit IRC
16:10:57 <jralls> looks like it also works to do `install_name_tool -add_rpath `pwd` libboost_locale.dylib`, confirming my thought that it's not ltdl's fault. That might not work as well for bundling, though.
16:12:09 <mohave> Okay, gnucash built and launches!
16:13:31 <jralls> \o/
16:21:09 *** User_ has quit IRC
16:21:51 *** jervin has joined #gnucash
16:23:24 *** Gerd has quit IRC
16:24:28 *** lmat has quit IRC
16:25:04 *** mohave has quit IRC
16:33:19 *** jervin has quit IRC
16:44:12 *** mohave has joined #gnucash
16:53:29 <mohave> jralls, do you know if gnucash needs to be run from a bundle for aqbanking to be available and for most of the app icons to appear?
16:53:52 *** omnireq has quit IRC
16:54:13 <mohave> some but not all icons do show when running from the install dir
16:55:14 *** omnireq has joined #gnucash
16:55:14 *** ChanServ sets mode: +v omnireq
16:57:52 *** KaiForce has quit IRC
16:59:44 *** frakturfreak has quit IRC
17:40:52 *** sbluhm has quit IRC
17:42:52 *** mohave has quit IRC
17:43:34 *** mohave has joined #gnucash
17:46:35 *** mohave has quit IRC
17:49:25 *** mohave has joined #gnucash
18:02:22 *** omnireq has quit IRC
18:02:32 *** omnireq has joined #gnucash
18:02:32 *** ChanServ sets mode: +v omnireq
18:11:25 *** phoenix has quit IRC
18:23:58 *** almean has joined #gnucash
18:40:50 *** almean has left #gnucash
18:59:55 *** waeking has quit IRC
19:00:06 *** waeking has joined #gnucash
19:00:06 *** ChanServ sets mode: +v waeking
19:00:26 *** Aussie_matt has joined #gnucash
19:24:26 *** mohave has quit IRC
19:24:47 *** mohave has joined #gnucash
19:27:48 *** mohave has quit IRC
19:51:40 <jralls> @tell mohave The icons require adwaita-icon-theme. That used to be in gtk-osx-themes.modules but I moved it and its two dependencies to gtk-osx.modules and added it to meta-gtk-osx-gtk3 yesterday so you can just rebuild that (and get a bunch of other updates as well).
19:51:40 <gncbot> jralls: The operation succeeded.
19:52:51 <jralls> @tell mohave I also changed the rpath-fix-script in the CMakeLists.txt for boost so that it changes the @rpath paths back to absolute paths.
19:52:51 <gncbot> jralls: The operation succeeded.
19:55:26 *** mohave has joined #gnucash
19:58:26 *** mohave has quit IRC
20:30:35 *** jervin has joined #gnucash
20:39:35 *** jervin has quit IRC
20:39:49 *** tonysoar has joined #gnucash
20:47:33 *** guak has quit IRC
20:48:26 *** jervin has joined #gnucash
21:00:27 *** cybertani[m] has joined #gnucash
21:10:09 <cybertani[m]> Hi, i'm new to gnucash. I install the project, use it a bit and look the website docs. I see that it isn't translated to spanish, so i would like to make a contribution, but inspecting the repo i was surprised that the english version has the "es" locale key.
21:10:10 <cybertani[m]> It's a bug? 🤔
21:10:10 <cybertani[m]> https://github.com/Gnucash/gnucash-htdocs/blob/master/po/es.po
21:15:51 *** tonysoar has quit IRC
21:21:15 <chris> cybertani[m]: you'll need to wait for some Europeans to be awake
21:29:14 *** chris has quit IRC
21:31:30 *** jervin has joined #gnucash
21:40:14 *** jervin has quit IRC
21:40:27 *** lmat has joined #gnucash
21:55:11 *** shoonya has joined #gnucash
22:08:24 *** ArtGravity has quit IRC
22:25:12 *** Mechtilde has joined #gnucash
22:30:24 *** Mechtilde has quit IRC
22:30:42 *** Mechtilde has joined #gnucash
22:41:49 *** aminbegood has joined #gnucash
22:43:48 *** mohave has joined #gnucash
22:48:00 *** mohave has quit IRC
23:04:31 *** shoonya has quit IRC
23:06:31 *** aminbegood has quit IRC
23:21:50 *** storyjesse has joined #gnucash
23:37:54 *** jervin has joined #gnucash
23:57:24 *** oedi has joined #gnucash