2020-05-30 GnuCash IRC logs

00:04:36 *** marusich has quit IRC
00:06:07 <SimpleIsGood> I am still having some difficulty with the mySQL save as procedure.
00:07:16 <SimpleIsGood> I'm running GnuCash version 3.10 and mySQL Workbench 8.0.11.
00:08:13 <SimpleIsGood> Here is the most recent Trace file after a Save As attempt:
00:08:13 <SimpleIsGood> * 00:03:28 WARN <gnc.app-utils> Could not spawn perl: Failed to execute child process (Bad file descriptor)
00:08:14 <SimpleIsGood> * 00:03:28 ERROR <gnc.app-utils> gnc_process_get_fd: assertion 'proc' failed
00:08:14 <SimpleIsGood> * 00:03:28 ERROR <gnc.app-utils> gnc_detach_process: assertion 'proc && proc->pid' failed
00:08:14 <SimpleIsGood> * 00:03:51 ERROR <gnc.backend.dbi> [GncDbiBackend<Type>::session_begin()] Unable to connect to database 'gnucash'
00:08:14 <SimpleIsGood> * 00:03:56 ERROR <gnc.engine.sx> gnc_sx_get_sxes_referencing_account: assertion 'sxactions != NULL' failed
00:09:52 <SimpleIsGood> I am able to "connect" to the database through MySQL Workbench. Any clues where to start would be much appreciated.
00:24:27 *** marusich has joined #gnucash
00:24:27 *** ChanServ sets mode: +v marusich
01:04:11 *** jervin1 has joined #gnucash
01:04:37 *** jervin has quit IRC
01:06:44 *** jervin has joined #gnucash
01:07:11 *** jervin1 has quit IRC
01:08:39 *** sbluhm has joined #gnucash
01:08:39 *** ChanServ sets mode: +v sbluhm
01:11:14 *** jervin has quit IRC
01:12:57 *** jervin has joined #gnucash
01:15:30 *** jervin1 has joined #gnucash
01:15:57 *** jervin has quit IRC
01:15:57 *** jervin1 is now known as jervin
01:19:44 *** Mechtilde has joined #gnucash
01:21:59 *** fell has quit IRC
01:23:19 *** fell has joined #gnucash
01:23:19 *** ChanServ sets mode: +o fell
01:36:23 *** sbluhm has quit IRC
01:38:57 *** jervin has quit IRC
01:39:52 *** Aussie_matt has joined #gnucash
01:45:07 *** storyjesse has joined #gnucash
01:51:36 <chris> @tell gjanssens jralls merged in running cli-reports. but saved-reports are not accessible: these must run load_system_config() and load_user_config().
01:51:36 <gncbot> chris: The operation succeeded.
02:05:07 *** storyjesse has quit IRC
02:33:25 *** gjanssens has joined #gnucash
02:33:25 *** ChanServ sets mode: +o gjanssens
02:40:43 <gjanssens> .
02:40:43 <gncbot> gjanssens: Sent 49 minutes ago: <chris> jralls merged in running cli-reports. but saved-reports are not accessible: these must run load_system_config() and load_user_config().
02:42:59 <gjanssens> chris: that's weird
02:43:20 <gjanssens> I had already merged those into my own branch and adapted to show reports.
02:43:30 <gjanssens> Looks like part of my work got lost somewhere :(
02:45:19 <gjanssens> Oh drat. I have been working on the wrong branch for the last few commits :(
02:45:26 <gjanssens> Let me fix that.
03:02:44 *** sbluhm has joined #gnucash
03:06:29 <gjanssens> The trouble is jralls made different fixes than I did, so rebasing will be challenging
03:12:40 *** storyjesse has joined #gnucash
03:31:58 *** sbluhm has quit IRC
03:34:15 <gjanssens> That should do it.
03:39:19 *** fabior has joined #gnucash
03:46:16 *** User_ has joined #gnucash
04:26:21 *** sbluhm has joined #gnucash
04:26:21 *** ChanServ sets mode: +v sbluhm
04:32:57 *** Aussie_matt has quit IRC
04:47:58 *** marusich has quit IRC
04:51:31 <chris> gjanssens: \o/
04:58:02 *** fabior has quit IRC
05:03:50 *** sbluhm has quit IRC
05:05:45 *** suukim has joined #gnucash
05:14:08 *** Mechtilde has quit IRC
05:16:20 <chris> gjanssens: last optional sanity check: add-price-quotes and run-report both don't check that datafile actually exists.
05:26:40 *** keiffer has joined #gnucash
05:27:00 *** fabior has joined #gnucash
05:32:59 *** Mechtilde has joined #gnucash
05:45:52 *** User_ has quit IRC
06:11:22 <chris> from the guile side, minr issue: if >1 report-name matches, it'll select one randomly. do we prevent this?
06:27:23 *** fabior has quit IRC
06:41:23 *** o01eg has quit IRC
06:44:33 <fell> gjanssens, some idea about https://lists.gnucash.org/logs/2020/05/30.html#T00:06:07 ?
06:56:25 *** finster has joined #gnucash
06:56:25 *** ChanServ sets mode: +v finster
07:09:42 *** lmat has quit IRC
07:12:44 *** lmat has joined #gnucash
07:18:57 *** lmat has quit IRC
07:20:33 *** lmat has joined #gnucash
07:39:59 *** Jimraehl1 has joined #gnucash
07:41:25 *** Jimraehl1 has quit IRC
07:43:06 *** User_ has joined #gnucash
08:59:08 *** Mechtilde has quit IRC
09:08:57 <fell> No Windows uploads today :-(
09:13:23 *** Mechtilde has joined #gnucash
09:17:08 *** Mechtilde has quit IRC
09:59:07 *** Mechtilde has joined #gnucash
10:25:08 *** Mechtilde has quit IRC
10:26:35 *** Mechtilde has joined #gnucash
10:56:47 *** jervin has joined #gnucash
11:00:54 *** jervin has quit IRC
11:20:08 <chris> "from the guile side, minr issue: if >1 report-name matches, it'll select one randomly. do we prevent this?" -> now multiple reportname matches are handled appropriately.
11:34:11 *** omnireq_ has joined #gnucash
11:35:21 *** omnireq has quit IRC
11:38:42 *** laerling has quit IRC
11:38:43 *** waeking has quit IRC
11:38:43 *** Trel has quit IRC
11:38:43 *** flips has quit IRC
11:38:43 *** Yotson has quit IRC
11:38:43 *** chris has quit IRC
11:38:43 *** Derperperd has quit IRC
11:38:48 *** Yotson has joined #gnucash
11:38:54 *** laerling has joined #gnucash
11:38:56 *** Trel has joined #gnucash
11:39:21 *** flips has joined #gnucash
11:39:22 *** ChanServ sets mode: +v flips
11:39:28 *** waeking has joined #gnucash
11:39:28 *** ChanServ sets mode: +v waeking
11:39:59 *** Derperperd has joined #gnucash
11:44:44 *** chris has joined #gnucash
11:44:44 *** ChanServ sets mode: +v chris
11:54:43 *** sbluhm has joined #gnucash
11:54:43 *** ChanServ sets mode: +v sbluhm
11:55:59 *** jervin has joined #gnucash
11:56:29 *** storyjesse has quit IRC
12:03:41 <jralls> gjanssens, the windows build failed for a missing function in gcc 9.3.0. Webkit finally finished, I'll try a build with gcc 10.1 now.
12:04:14 <jralls> gjanssens, the function in question is a deep requirement from boost::program_options.
12:14:56 <chris> both 3.11 and 3.903 are looking good
12:29:12 *** sbluhm has quit IRC
12:38:08 <jralls> gjanssens, no joy. gnucash.cpp fails because /mingw32/include/c++/10.1.0/cmath is broken. gnucash-core-app.cpp also /mingw32/include/c++/10.1.0/bits/basic_string.h, vsprintf and vsnprintf.
12:38:24 <jralls> chris, No, master is *not* looking good.
13:16:36 *** jervin has quit IRC
14:02:32 *** sbluhm has joined #gnucash
14:02:32 *** ChanServ sets mode: +v sbluhm
14:12:08 *** suukim has quit IRC
14:13:47 *** sbluhm has quit IRC
14:40:05 *** frakturfreak has joined #gnucash
14:40:05 *** ChanServ sets mode: +v frakturfreak
15:05:36 *** omnireq_ has quit IRC
15:05:50 *** omnireq has joined #gnucash
15:27:01 *** sbluhm has joined #gnucash
15:27:01 *** ChanServ sets mode: +v sbluhm
15:33:09 <gjanssens> jralls: that sucks royally :(
15:35:22 <gjanssens> jralls: what do you propose now ?
15:35:30 <jralls> gjanssens, there are also a bunch of function not found errors for the windows locale code.
15:37:28 <jralls> That's a difficult question to answer. The quickest fix would be to revert the boost::program_options and boost::locale commits and hope that that gets it to build.
15:38:35 <jralls> The better one is to try and fix the failures, but that's going to take a few days.
15:39:11 <jralls> Well, maybe only a day or two, but the release goes to Wednesday at the earliest.
15:41:57 <gjanssens> Personally I'd vote for the second option.
15:42:05 <gjanssens> (obviously)
15:42:33 <gjanssens> These are beta releases. It's not the end of the world if one slips a few days
15:43:01 <gjanssens> I *am* surprised though the boost::locale commits cause issues.
15:43:19 <gjanssens> We have been using that in the 3.x series as well.
15:44:44 <jralls> That might not be boost::locale, it might be moving the windows localization functions to a C++ file.
15:45:04 <gjanssens> Oh, indeed.
15:47:43 <jralls> This one puzzles me, though: 'libtintl_vsnprintf' is not a member of 'std'; did you mean 'libintl_vsnprintf'?
15:48:21 <jralls> from ming32/includ/c++/10.1.0/bits/basic_string.h:6652:50
15:48:30 <gjanssens> As for boost::program_options, I briefly thought to try another method of storing values than value_semantics (the library also has support for boost::optional), however it already fails at header includes so that would not change anything.
15:49:31 *** omnireq_ has joined #gnucash
15:49:32 <gjanssens> was that copy/pasted ? The first has an additional 't' after 'lib'.
15:50:19 <jralls> No, vmtools isn't working today so I can't copy/paste between windows and mac
15:50:43 *** omnireq has quit IRC
15:50:55 <jralls> hence includ instead of include in the second line
15:52:02 <gjanssens> So the 't' in there was a typo on your behalf ?
15:52:38 *** Mechtilde has quit IRC
15:53:10 <jralls> Yeah. "libtint" would have something to do with colors, but I dunno what libtintl would do. :-/
15:54:06 <gjanssens> Is current gnucash-on-windows in a state I can retry your build ? Or do you have that log somewhere (using gcc 10.1) ?
15:54:46 *** sbluhm has quit IRC
15:55:37 <jralls> I don't think it matters, it broke last night on code VM with gcc 9.3. https://code.gnucash.org/builds/win32/build-logs/master/2020-05/build-master-2020-05-30-03-02-40.log
15:56:14 <gjanssens> True, but I didn't see the errors there about libintl_vnsprintf ?
15:56:22 <gjanssens> So those seem to be new in 10.1
15:57:14 <jralls> Not necessarily. I think I have more cores on my VM than code does.
15:57:58 <jralls> I'll go comment out the offending line on code and start a build...
16:00:36 <jralls> Started.
16:02:28 <gjanssens> Ok
16:06:09 <jralls> Meanwhile, is your VM in a buildable state or does it need the new webkit?
16:06:21 <jralls> Which is up on SF, btw.
16:07:15 <gjanssens> jralls: my windows build continues to fail on guile2. I have just removed both source and build dir for guile and restarted the build
16:07:47 <gjanssens> Ok this time it did apply the patches, so that should get me a bit further.
16:08:39 <jralls> There's a new one of that on SF as well, has the patches already applied:https://sourceforge.net/projects/gnucash/files/Dependencies/guile-2.2.7.4-9d7759.tar.xz
16:09:41 <jralls> I just haven't pushed the setup_mingw64.ps1 and gnucash.modules changes yet.
16:10:37 <jralls> Oh! I forgot about adding gnucash-cli to gnucash.iss, too...
16:11:11 *** sbluhm has joined #gnucash
16:11:11 *** ChanServ sets mode: +v sbluhm
16:12:40 <jralls> Heh, no need. That grabs everything in @INST_DIR@\bin.
16:17:17 <jralls> gjanssens: https://code.gnucash.org/builds/win32/build-logs/master/2020-05/build-master-2020-05-30-16-00-08.log
16:17:19 <gjanssens> That's easy :)
16:17:32 <gjanssens> And my guile build still fails in exactly the same way...
16:18:20 <jralls> Try the new one?
16:20:11 <gjanssens> Ok.
16:20:54 <gjanssens> As for the vsnprintf weirdness, perhaps we should include gi18n.h inside the extern "C" block.
16:20:56 <gjanssens> Can you try that ?
16:22:36 <gjanssens> It may not play well when interpreted in a C++ context
16:22:50 <gjanssens> jralls: ^
16:24:44 <jralls> That worked, leaving just the windows localization functions.
16:25:52 <gjanssens> jralls: perhaps the Windows.h file needs the same treatment ?
16:26:03 <jralls> I tried that.
16:26:36 <gjanssens> I actually expected you to...
16:27:03 <gjanssens> The errors remain the same ?
16:28:04 <jralls> Yeah, no change. But I guess that's not surprising, it's "symbol not declared" at compile time not "symbol not found" at link time.
16:28:17 <gjanssens> True.
16:28:39 <gjanssens> I did remove quite a bit of includes at some point. I suspect one of those was needed for Windows.
16:30:06 <gjanssens> Perhaps stdio.h or stdlib.h ?
16:32:20 <jralls> No, it's Microsoft symbols that aren't found.
16:32:33 <jralls> They should be pulled in from Windows.h.
16:38:38 <gjanssens> IsValidLocaleName is defined in winnls.h, which is included in windows.h if RC_INVOKED is defined and NOWINRES isn't
16:40:45 <gjanssens> Eh no the other way around. If either RC_INVOKED is not defined or NOWINRES is
16:43:26 *** sbluhm has quit IRC
16:45:35 <jralls> I went for belt-and-suspeders: #ifdef RC_INVOKED #undef RC_INVOKED #endif #define NOWINRES 1. Building now.
16:46:23 <jralls> And no joy.
16:46:51 <gjanssens> Hmm, there's a second windows.h file in boost includes. Would that be interfering ?
16:47:36 <jralls> Shouldn't as long as it has compile guards.
16:48:12 <gjanssens> And it's not in the include paths either so that would be odd
16:48:43 <jralls> But maybe it would be easier to just have gnucash-windows-locale.c and gnucash-mac-locale.m.
16:49:07 <jralls> The latter so your editor doesn't confuse objective C with C++.
16:49:33 <gjanssens> Yeah, that would probably work around this issue indeed.
16:50:22 <gjanssens> Also my guile2 is still bailing out, even with your newer version. So I now think my environment is corrupt after all.
16:50:30 <gjanssens> Will have to restart from scratch
16:50:54 <jralls> :(
16:51:00 <gjanssens> But that will be for later. As it's getting late, I'll have to leave you with it for now.
16:51:14 <gjanssens> Good luck and see you later!
16:51:18 <jralls> In which case you'll definitely need the new setup_mingw64.
16:51:33 <jralls> So I'll go ahead and push.
16:51:36 <gjanssens> Yeah, and probably even a Windows 10 VM
16:51:39 <jralls> Good night!
16:51:48 <gjanssens> I have the install usb ready to go.
16:51:51 <gjanssens> Bye!
16:52:28 *** gjanssens has quit IRC
17:27:36 <jralls> @tell gjanssens Moving set_win32_thread_locale() to gnucash-windows-locale() worked as a workaround. I pushed that and the relocation of gi18n.h to C linkage land. That leaves the change to include/c++/cmath.
17:27:36 <gncbot> jralls: The operation succeeded.
17:29:12 <jralls> @tell gjanssens I also pushed updates to gnucash.modules and setup_mingw32.ps1 for the new guile and webkit.
17:29:12 <gncbot> jralls: The operation succeeded.
18:20:46 *** SimpleIsGood1 has joined #gnucash
18:22:09 *** SimpleIsGood has quit IRC
18:24:15 *** frakturfreak has quit IRC
18:35:58 <jralls> fell: The TP Ukrainian translator has gotten ahead of you, only 12 untranslated strings and no fuzzies on master! ;-)
18:36:24 <fell> Cool!
18:38:14 *** frakturfreak has joined #gnucash
18:39:51 <fell> I am listenig this weekend to https://moers-festival.de/
18:40:38 <fell> WhenI was young it was open air, this year it free streamed.
18:47:05 *** SimpleIsGood has joined #gnucash
18:47:46 *** SimpleIsGood1 has quit IRC
18:52:46 *** User_ has quit IRC
18:55:27 <jralls> There's a lot of that this year. I "attended" my annual genealogy conference that way last week.
19:03:18 *** omnireq_ has quit IRC
19:03:30 *** omnireq_ has joined #gnucash
19:07:55 <jralls> chris, is gnucash-cli run-reports completely broken or can it run some things, and if the latter, what?
19:08:46 *** frakturfreak has quit IRC
19:42:16 *** SimpleIsGood1 has joined #gnucash
19:43:05 *** SimpleIsGood has quit IRC
20:03:28 *** SimpleIsGood1 has left #gnucash
20:21:03 *** omnireq__ has joined #gnucash
20:22:13 *** omnireq_ has quit IRC
20:35:07 *** jralls has quit IRC
20:41:24 *** jralls has joined #gnucash
20:41:25 *** ChanServ sets mode: +o jralls
21:00:23 *** jervin has joined #gnucash
21:02:59 *** marusich has joined #gnucash
21:02:59 *** ChanServ sets mode: +v marusich
21:13:36 *** keiffer has quit IRC
21:26:18 *** omnireq__ has quit IRC
21:26:38 *** omnireq__ has joined #gnucash
21:42:03 *** TownsendHardware has quit IRC
21:43:28 <chris> jralls: from my POV cli-reports are working very well
21:47:08 *** chris has quit IRC
21:47:46 *** chris has joined #gnucash
21:47:46 *** ChanServ sets mode: +v chris
21:49:08 <chris> jralls: have you completed c++options read-options? would it help to convert the saved-reports arbitrary eval-string to a custom scheme reader in the meantime?
21:54:36 <chris> cli-reports have 3 major syntaxes:
21:55:16 <chris> $gnucash --run-report=[reportname/guid] datafile.gnucash
21:55:31 <chris> $gnucash --run-report=[reportname/guid] --output-file=x.html datafile.gnucash
21:55:44 <chris> $gnucash --run-report=[reportname/guid] --output-file=x.html --export-type=TYPE datafile.gnucash
22:09:09 <fell> jralls, I just saw opensuse stopped shipping gnucash-docs after you moved them in the program direcory instead of the gnucash-docs directory.
22:18:02 <chris> fell: I've left the report-core.scm cli messages untranslated. not sure it's important to _ all of them.
22:21:36 <chris> fell: if the report-core.scm messages can be improved please do so
22:26:40 <fell> chris remind me in afew days. I will wait for some stabilization.
22:37:36 *** Gerd1 has joined #gnucash
22:38:12 *** Gerd has quit IRC
22:38:12 *** Gerd1 is now known as Gerd
22:48:57 *** jervin has quit IRC
23:47:48 *** jervin has joined #gnucash