2021-07-02 GnuCash IRC logs

00:56:21 *** Pegasus_RPG has quit IRC
01:12:49 *** fell has quit IRC
01:14:08 *** fell has joined #gnucash
01:14:08 *** ChanServ sets mode: +o fell
01:18:33 *** Pegasus_RPG has joined #gnucash
01:28:20 *** sbluhm has joined #gnucash
01:43:11 *** frakturfreak has quit IRC
01:50:18 *** sbluhm has quit IRC
01:51:31 *** sbluhm has joined #gnucash
01:55:14 *** Pegasus_RPG has quit IRC
01:57:12 *** frakturfreak has joined #gnucash
02:53:40 <Simon> Except that I don't use the import features in gnucash
02:54:42 <Simon> I have two scripts that add transactions; one is using exact times to the second from the bank and the other increments date-entered by 1s for each transaction
02:59:06 *** Mechtilde has quit IRC
03:01:34 *** field^Zzz3 has joined #gnucash
03:16:12 *** r2d3 has joined #gnucash
03:16:26 *** gjanssens has joined #gnucash
03:16:26 *** ChanServ sets mode: +o gjanssens
03:16:35 <gjanssens> .
03:16:35 <gncbot> gjanssens: Sent 11 hours and 20 minutes ago: <warlord> I don't think that was mirrored, but I guess we'll figure that out shortly.
03:16:43 *** gjanssens sets mode: +o gncbot
03:19:42 *** field^Zzz3 has quit IRC
03:29:08 *** fell has quit IRC
03:29:29 *** fell has joined #gnucash
03:29:29 *** ChanServ sets mode: +o fell
03:53:26 *** field^Zzz3 has joined #gnucash
04:01:44 *** fell has quit IRC
04:04:46 *** fell has joined #gnucash
04:04:46 *** ChanServ sets mode: +o fell
04:08:24 *** r2d3 has quit IRC
04:08:38 *** r2d3 has joined #gnucash
04:12:56 *** r2d3 has quit IRC
04:13:09 *** r2d3 has joined #gnucash
04:13:37 *** Mechtilde has joined #gnucash
04:33:57 *** r2d3 has quit IRC
04:34:40 <Simon> two of these transaction pairs happened in the same second, which is why they have the same times
04:36:16 <Simon> one of them has a side with 5 splits, maybe it's possible to get the same date-entered by adding/removing splits to get two separate transactions?
04:36:31 <Simon> although that shouldn't be possible
04:41:24 <Simon> for the remaining two there is no credible reason for why they would have the same date-entered
04:42:07 <Simon> although it was in 2006 when I was first entering transactions so it's not impossible that I imported them
04:43:22 <Simon> for the 157 transactions that have a date-entered of 2008-03-17 19:08:46... these are scheduled transactions
04:45:25 <Simon> the last one has a date-posted of 2021-09-24 and would have been created on the 26th of June
04:52:49 <Simon> I expect to have more scheduled transactions created in 2-3 days so I can check what date-entered they have
05:15:18 *** fell has quit IRC
05:31:15 *** Mechtilde has quit IRC
05:32:25 *** Mechtilde has joined #gnucash
05:38:06 *** fell has joined #gnucash
05:38:06 *** ChanServ sets mode: +o fell
05:50:03 *** AstronautSurfer has joined #gnucash
05:50:03 *** ChanServ sets mode: +v AstronautSurfer
05:56:46 *** r2d3 has joined #gnucash
05:59:04 *** AstronautSurfer has quit IRC
06:00:04 *** r2d3 has quit IRC
06:00:17 *** r2d3 has joined #gnucash
06:02:36 *** User has joined #gnucash
06:05:44 *** bertbob has quit IRC
06:10:03 *** bertbob has joined #gnucash
06:10:04 *** ChanServ sets mode: +v bertbob
06:10:33 *** r2d3 has quit IRC
06:10:46 *** r2d3 has joined #gnucash
06:11:46 *** sbluhm has quit IRC
06:15:05 *** r2d3 has quit IRC
06:15:18 *** r2d3 has joined #gnucash
06:29:51 *** sbluhm has joined #gnucash
06:48:50 *** celeste has quit IRC
06:57:09 *** celeste has joined #gnucash
06:57:09 *** ChanServ sets mode: +v celeste
07:25:31 *** r2d3 has quit IRC
07:27:45 *** r2d3 has joined #gnucash
07:31:37 *** r2d3 has quit IRC
07:57:36 *** sbluhm has quit IRC
08:17:26 <warlord> Simon, did you create an SX that fired off 157 instances at once? That would also "do it".
08:17:36 <warlord> jralls, yeah, didn't quite think about importing..
09:16:12 *** Aussie_matt has quit IRC
09:29:13 *** Jimraehl1 has joined #gnucash
09:31:05 *** Jimraehl1 has left #gnucash
09:32:08 *** Jimraehl1 has joined #gnucash
09:33:09 *** Jimraehl1 has left #gnucash
09:34:47 *** Jimraehl1 has joined #gnucash
09:35:29 *** Jimraehl1 has quit IRC
09:37:10 *** raeburn has quit IRC
10:00:07 *** AstronautSurfer has joined #gnucash
10:00:07 *** ChanServ sets mode: +v AstronautSurfer
10:40:34 *** Pegasus_RPG has joined #gnucash
10:43:54 <Simon> warlord: no, this is a monthly SX and it (v2.6) has been reusing the date-entered of the SX template transaction
10:44:23 <Simon> not sure if 4.6 has created any SX yet but when it does I can check the date-entered on them
10:45:05 <warlord> Oh.. Hmm... That seems like a bug to me!
10:45:53 <warlord> Although the date-posted should be different, and that takes precedence..
10:46:32 <Simon> except when I have more than one of them for any reason
10:46:35 *** Pegasus_RPG has quit IRC
10:46:47 *** Mechtilde has quit IRC
10:46:53 <Simon> it should really use the current time
10:49:11 <Simon> and probably increment it by 1s for each transaction in the same SX
10:51:24 <Simon> 4.6 is doing it too
10:51:50 <Simon> I have a new transaction with date-entered in 2014 because that's the SX template
10:53:46 <warlord> please file a bug on that.
10:58:23 <Simon> done; https://bugs.gnucash.org/show_bug.cgi?id=798224
10:58:41 <warlord> thx
11:11:19 *** storyjesse has quit IRC
11:15:25 *** ArtGravity has joined #gnucash
11:15:25 *** ChanServ sets mode: +v ArtGravity
11:36:31 *** mydogsnameisrudy has joined #gnucash
11:39:27 *** mydogsnameisrudy has quit IRC
11:44:00 <gjanssens> chris: looks like your guile modularization effort has left us with several circular dependencies in guile sources. Those cause issues while trying to compile from scratch.
11:45:49 <gjanssens> The ones I ran into today for example: (gnucash report commodity-utilities) depends on (gnucash report report-utilities) depends on (gnucash report html-acct-table) depends on (gnucash report commodity-utilities)
11:46:37 <gjanssens> And (gnucash report report-utilities) depends on (gnucash report html-acct-table) depends on (gnucash report report-utilities)
11:47:25 <gjanssens> You can't do that as guile will choke on those in a clean build directory, particularly if the build concurrency is 1 (that is, only one file is built at once).
11:48:14 <gjanssens> Same issue in the qif importer
11:48:53 <chris> ugh. what's the cli commands to highlight these errors?
11:52:02 <gjanssens> In my case it was running ninja-build in a clean working directory.
11:52:32 <gjanssens> It got stuck at about 202 files left to build.
11:54:09 <gjanssens> Where it gets weird is that if I run a 'cmake .' in the same directory after the build failure, suddenly the next build does work (after 'cmake .' there were about 320 files to rebuild)
11:54:43 <chris> odd that the 1000 odd files have never failed to build for me
11:56:13 <gjanssens> Right, perhaps I'm jumping to conclusions.
11:56:53 <chris> I think you're right though. these 3-level module names are not typically imported, so, an easy hack is to combine them, at the cost of less fast recompilation
11:57:20 <gjanssens> After the successful build I did another 'ninja-build clean; ninja-build' and this time it worked flawlessly. Go figure.
11:57:43 <gjanssens> Which brings me to what happened right before the build attempt.
11:58:53 <gjanssens> That is, I pulled most recent maint from the upstreram gnucash git repo, built it (by running ninja-build) and when executing the freshly built gnucash most reports were missing.
11:59:11 <chris> now, reports are expected to (use-module (gnucash report)) rather than (gnucash report commodity-utilities), so, I guess it's safe to move functions around
11:59:58 <gjanssens> That's a frequently recurring pattern on my system and I have not found the cause yet.
12:00:16 <gjanssens> So after that I do a ninja-build clean; ninja-build
12:00:47 <chris> (IIRC your ninja-build = my ninja)
12:00:51 <chris> (IIUC your ninja-build = my ninja)
12:02:55 <chris> Another way to fix these dependencies is to remove (use-modules) but allow some code duplication e.g. small utility functions
12:03:06 <gjanssens> Which only removes about 858 files rather than 962. That is weird. So I'll assume these files that didn't get removed are breaking things.
12:03:33 <gjanssens> You understand right. The name of the tool is distro dependend
12:03:58 <chris> I typically just trash the build dir
12:03:59 <gjanssens> But for the life of me I don't understand why that happens from time to time.
12:04:50 <gjanssens> I prefer to save time by not doing so. And it usually works fine, except sometimes it doesn't.
12:05:39 <chris> gtg now....
12:05:51 <gjanssens> Anyway, the oddity on my side is clearly not a reason to reorganize files. Perhaps my tax fix from a few days back was reduntant as well
12:05:53 <gjanssens> Bye!
12:29:31 *** guak has joined #gnucash
12:50:35 <jralls> gjanssens I've run into these problems and complained to chris about circular dependencies before, but the problem pre-dates him. The hackish work-around is the symlinked .scm files in the builddir. That works because guile can resolve its symbols from them as well as from the .go files.
12:51:46 <gjanssens> jralls: ok. I brought it up with chris as I traced these changes back to his modularity work. But apparently there's something else odd going on.
12:52:59 <gjanssens> Probably your suggestion to the symlinks is the ugly but real solution.
12:53:04 <jralls> Yeah, it has to do with making and removing those symlinks. Sometimes ninja clean removes them but then ninja doesn't put them back. Running cmake again fixes it somehow.
12:53:46 <jralls> It's not a suggestion, it's been there for at least 10 years, maybe longer.
12:54:08 <jralls> As for ugly, definitely.
12:54:18 <gjanssens> Right.
12:55:22 <gjanssens> Got to bo again, though the hint at symlinks/ninja-build/cmake is an interesting one. I may try to dig deeper into that bit later.
12:55:31 <gjanssens> s/bo/go/ ...
12:55:41 <jralls> ;-)
12:56:07 <jralls> The simplest way to avoid the problem is rm -rf * && cmake && ninja.
12:57:25 <jralls> You only need that when switching branches or after a big pull. There's another reason to do it too: Swig files also have dependency tracking problems.
13:01:06 *** sbluhm has joined #gnucash
13:19:54 *** Mechtilde has joined #gnucash
13:20:36 *** sbluhm has quit IRC
13:38:37 *** sbluhm has joined #gnucash
13:39:25 *** AstronautSurfer has quit IRC
13:39:30 *** Hamaryns has joined #gnucash
13:39:30 *** ChanServ sets mode: +v Hamaryns
13:52:52 *** sbluhm has quit IRC
14:01:34 *** sbluhm has joined #gnucash
14:02:07 *** kcin has joined #gnucash
14:11:42 *** kcin1 has joined #gnucash
14:13:37 *** kcin has quit IRC
14:14:10 *** kcin1 is now known as kcin
14:35:53 *** Jimraehl1 has joined #gnucash
14:36:32 *** Jimraehl1 has quit IRC
14:58:09 <Simon> this fixes the Ambiance theme's rounded inset border on the register cells: .gnc-class-register-foreground { border: none; background-image: none; border-radius: 0; }
14:58:34 <Simon> and this finally makes it make possible for me to see the text on account tabs with dark colours...: notebook box label { color: black; background-color: rgba(255,255,255,0.35); }
15:39:50 *** celeste has quit IRC
15:50:48 *** Hamaryns has quit IRC
16:00:49 *** sbluhm has quit IRC
16:32:40 *** field^Zzz3 has quit IRC
16:33:28 *** kcin1 has joined #gnucash
16:33:52 *** kcin has quit IRC
16:34:54 *** Mechtilde has quit IRC
16:35:56 *** kcin1 is now known as kcin
17:07:30 *** Hamaryns has joined #gnucash
17:07:30 *** ChanServ sets mode: +v Hamaryns
17:08:08 *** Hamaryns has quit IRC
17:18:18 *** gjanssens has quit IRC
17:25:32 *** kcin has quit IRC
17:33:43 *** User has quit IRC
17:40:45 *** celeste has joined #gnucash
17:40:45 *** ChanServ sets mode: +v celeste
17:41:15 <jralls> 4
18:15:22 *** David1 has joined #gnucash
18:48:44 *** Pegasus_RPG has joined #gnucash
18:59:48 *** jervin has quit IRC
19:22:04 *** Pegasus_RPG has quit IRC
19:32:15 *** Pegasus_RPG has joined #gnucash
19:33:16 *** jervin has joined #gnucash
19:39:26 *** ArtGravity has quit IRC
19:49:27 *** Pegasus_RPG has quit IRC
20:13:50 *** guak has quit IRC
20:35:41 *** celeste has quit IRC
21:17:47 *** celeste has joined #gnucash
21:17:47 *** ChanServ sets mode: +v celeste
21:41:48 *** User has joined #gnucash
22:16:28 *** User has quit IRC
22:18:26 *** storyjesse has joined #gnucash
22:25:23 *** Aussie_matt has joined #gnucash
22:46:21 <chris> @tell gjanssens do test 797596 - it's also in my branch bug797596 and the payment->payment analysis is still "unstyled"
22:46:21 <gncbot> chris: The operation succeeded.
23:06:13 *** storyjesse has quit IRC