2020-02-23 GnuCash IRC logs
00:48:08 *** omnireq_ has quit IRC
00:50:40 *** omnireq has joined #gnucash
00:50:40 *** ChanServ sets mode: +v omnireq
01:09:39 *** Gerd has quit IRC
01:11:13 *** omnireq has quit IRC
01:11:25 *** omnireq has joined #gnucash
01:11:25 *** ChanServ sets mode: +v omnireq
01:13:25 *** omnireq_ has joined #gnucash
01:14:39 *** omnireq has quit IRC
01:26:38 *** Gerd has joined #gnucash
01:26:44 *** Gerd1 has joined #gnucash
01:29:38 *** Gerd has quit IRC
01:29:38 *** Gerd1 is now known as Gerd
01:30:48 *** bertbob has quit IRC
01:40:44 *** fell has quit IRC
01:42:04 *** fell has joined #gnucash
01:42:04 *** ChanServ sets mode: +o fell
01:52:43 *** omnireq_ has quit IRC
01:53:59 *** omnireq_ has joined #gnucash
01:59:28 *** sbluhm has joined #gnucash
01:59:28 *** ChanServ sets mode: +v sbluhm
02:15:13 *** sbluhm has quit IRC
02:17:17 *** JayC has joined #gnucash
02:17:18 *** ChanServ sets mode: +v JayC
02:27:35 *** Gerd has quit IRC
02:49:46 *** Gerd has joined #gnucash
03:10:10 *** sbluhm has joined #gnucash
03:10:10 *** ChanServ sets mode: +v sbluhm
03:39:25 *** sbluhm has quit IRC
03:40:54 *** sbluhm has joined #gnucash
03:47:45 *** Gerd has quit IRC
03:52:09 *** sbluhm has quit IRC
04:00:28 *** apteryx_ has joined #gnucash
04:02:16 *** apteryx has quit IRC
04:03:00 *** sbluhm has joined #gnucash
04:03:00 *** ChanServ sets mode: +v sbluhm
04:06:40 *** Gerd has joined #gnucash
04:07:21 *** bertbob has joined #gnucash
04:07:22 *** ChanServ sets mode: +v bertbob
04:09:46 *** sbluhm has quit IRC
04:15:58 *** sbluhm has joined #gnucash
04:15:58 *** ChanServ sets mode: +v sbluhm
04:29:10 *** Gerd has quit IRC
04:31:43 *** sbluhm has quit IRC
05:04:24 *** finster has quit IRC
05:06:17 *** Gerd has joined #gnucash
05:09:50 *** finster has joined #gnucash
05:09:50 *** ChanServ sets mode: +v finster
05:16:13 *** omnireq_ has quit IRC
05:16:23 *** omnireq_ has joined #gnucash
05:19:42 *** blathijs has quit IRC
05:28:47 *** Gerd has quit IRC
05:36:18 *** blathijs has joined #gnucash
05:38:48 *** sbluhm has joined #gnucash
05:38:48 *** ChanServ sets mode: +v sbluhm
05:43:56 *** Aussie_matt has joined #gnucash
05:46:18 *** blathijs has quit IRC
05:52:18 *** sbluhm has quit IRC
05:58:31 *** blathijs has joined #gnucash
06:18:37 *** gjanssens has joined #gnucash
06:18:38 *** gncbot sets mode: +o gjanssens
06:18:38 *** ChanServ sets mode: +o gjanssens
06:18:44 <gjanssens> .
06:18:44 <gncbot> gjanssens: Sent 9 hours and 43 minutes ago: <jralls> I've fixed all of the issues you raised so far on c++options and force-pushed the result, losing all of your comments. The suggestion about gnc_account_list_from_types turned out to be a huge change and I've decided to set it aside for now.
06:18:45 <gncbot> gjanssens: Sent 9 hours and 40 minutes ago: <jralls> I found a better way to flag _get_option(): I converted it to a friend function swig_get_option implemented in gnc-optiondb.i. The friend decl still has to be public, but it's pretty clear now that it's intended only for swig.
06:18:46 <gncbot> gjanssens: Sent 9 hours and 35 minutes ago: <jralls> It builds on a more-or-less up to date Arch Linux but the Integrate gnc-option into app-utils commit causes some odd test failures in test-sx, test-aqb, and python-bindings. Still trying to figure those out.
06:20:08 <gjanssens> jralls: nice and ok on the gnc_account_list_from_types
06:20:48 <gjanssens> jralls: we can always do that later when other code is ready to reuse it.
06:22:41 *** sbluhm has joined #gnucash
06:22:41 *** ChanServ sets mode: +v sbluhm
06:24:42 <gjanssens> jralls: the friend function is an elegant solution. Good work
06:35:40 *** oozer has joined #gnucash
06:36:19 *** Aussie_matt has quit IRC
07:24:53 *** Gerd has joined #gnucash
07:26:47 <fell> gjanssens: *** Error during phase clean of gnucash-docs-git: ########## Error running ninja clean *** [15/16] :-(
07:29:35 <gjanssens> fell: ok, this sucks
07:40:02 *** puck has quit IRC
07:42:16 *** puck has joined #gnucash
07:45:57 *** sbluhm has quit IRC
08:16:04 *** Gerd has quit IRC
08:38:13 *** omnireq_ has quit IRC
08:40:35 *** omnireq_ has joined #gnucash
08:57:25 *** sbluhm has joined #gnucash
08:57:25 *** ChanServ sets mode: +v sbluhm
09:00:13 *** omnireq_ has quit IRC
09:01:28 *** omnireq_ has joined #gnucash
10:06:14 *** chris has quit IRC
10:08:47 *** chris has joined #gnucash
10:08:47 *** ChanServ sets mode: +v chris
10:59:37 *** warlord2 has joined #gnucash
11:00:52 *** warlord has quit IRC
11:04:32 *** TownsendHardware has joined #gnucash
11:09:25 *** sbluhm has quit IRC
11:16:54 *** jervin has joined #gnucash
11:23:17 *** jervin has quit IRC
11:23:38 *** jervin has joined #gnucash
11:27:58 *** KevinDB has quit IRC
11:29:13 *** KevinDB has joined #gnucash
11:29:13 *** ChanServ sets mode: +v KevinDB
11:34:23 *** Gerd has joined #gnucash
12:14:42 *** jervin has quit IRC
12:18:06 <gjanssens> fell: seems to be a ninja bug/weakness https://github.com/ninja-build/ninja/pull/1334
12:18:28 <gjanssens> If you can't track all generated files (as is the case in our documentation) ninja's clean command fails
12:19:00 <gjanssens> make doesn't have this issue, so I will switch to cmake/make combo for windows
12:20:10 <fell> ... with a comment containing above ref.
12:32:21 <gjanssens> Done. Let's see what tomorrow's build will do - and more interestingly the day after tomorrow's build
12:33:37 <jralls> gjanssens, could you have a look at https://bugs.gnucash.org/show_bug.cgi?id=797621?
12:42:43 *** omnireq_ has quit IRC
12:43:07 *** sbluhm has joined #gnucash
12:43:59 *** omnireq_ has joined #gnucash
12:47:28 *** Gerd has quit IRC
12:51:23 <jralls> gjanssens, I found the Arch Linux problem, fixed it, and force-pushed so I hope everything will build and test cleanly for you too.
12:52:52 *** sbluhm has quit IRC
12:58:24 <jralls> The problem was that I had shadowed qof_book_new in gnc-optiondb.i--it's not wrapped in engine.i--and the three tests were picking up my shadow instead of the real one. It seems that for app-utils the swig bindings are being built into libgnc-app-utils.so instead of gnucash/_sw_app_utils.so. Is that intentional?
13:12:39 <gjanssens> jralls: not by my intention at least
13:13:22 <gjanssens> The end goal is to move all binding stuff to bindings anyway, but that requires some more decoupling in app-utils
13:20:21 <jralls> A lot more, I imagine.
13:21:16 <jralls> Since book uses options (and will even more if we move saved reports into the book file) should options move over to engine as part of this exercise?
13:22:51 <gjanssens> In that light it would indeed make sense to promote it engine indeed
13:23:56 <gjanssens> Though I doubt you can do that without untangling (disentaginling?) the whole scheme-c spaghetti in app-utils
13:24:12 <gjanssens> I don't know if options is sufficiently isolated in there.
13:25:13 *** omnireq_ has quit IRC
13:26:03 <jralls> It's not, but getting rid of it is the whole point of c++options.
13:26:13 <gjanssens> Of course
13:26:29 <gjanssens> So there's hope :)
13:26:47 <jralls> It's turned out to be a complete rewrite which is why it's taking so long.
13:27:02 <gjanssens> Re the bug you asked about, what input can I give ?
13:27:21 <gjanssens> So far there doesn't seem to be much to point at the source of the issue ?
13:27:33 <gjanssens> I tried on my test book, but I don't experience any delays
13:27:34 *** omnireq_ has joined #gnucash
13:27:49 <gjanssens> That's on Fedora 31, current maint build
13:29:48 <jralls> So far it would appear to be a query from the invoice dialog blowing up. I've never spent any time at all with the business code so I don't know why it would need a query to fill in a popup menu.
13:30:32 <jralls> Most of our other popups are simple display the objects in qof-collection x.
13:44:11 *** Han has quit IRC
13:46:43 *** omnireq_ has quit IRC
13:48:23 *** omnireq has joined #gnucash
13:48:23 *** ChanServ sets mode: +v omnireq
13:51:06 *** Han has joined #gnucash
13:58:48 *** ramontjunior has joined #gnucash
14:02:52 <gjanssens> The windows log really only seems to show loading the book.
14:03:07 <gjanssens> I am looking at the entry ledger code now.
14:03:53 <gjanssens> Unfortunately I have always found register related code pretty opaque and hard to penetrate :(
14:04:20 <jralls> Yeah, another major cleanup job in our future. :-(
14:05:12 *** sbluhm has joined #gnucash
14:05:12 *** ChanServ sets mode: +v sbluhm
14:05:29 <gjanssens> So far I have found that refreshing the ledger runs plenty of queries, https://github.com/Gnucash/gnucash/blob/maint/gnucash/register/ledger-core/gncEntryLedgerLoad.c#L326
14:06:17 <gjanssens> But I don't even know yet if that code gets called when clicking on Taxable?, TaxIncluded or TaxTable fields in the ledger
14:06:51 <gjanssens> It is called when refreshing the register, so I'd expect the delays to be there.
14:06:56 <gjanssens> Looking further...
14:07:28 <gjanssens> FTR On my Windows test box I don't experience any slowdowns either. Perhaps my book is too small ?
14:07:37 <gjanssens> Or my VM too fast ?
14:08:42 <jralls> Or you have a reasonable amount of memory. Page faults are a serious performance killer, and he does report a memory usage problem.
14:08:59 <gjanssens> Indeed that's possible as well.
14:10:09 <gjanssens> The Windows VM has 4Gigs, my linux machine 16
14:10:58 <gjanssens> Windows is still 7 though (my Win10 install disk is waiting to be installed but not very high on my priority list...)
14:12:14 *** jervin has joined #gnucash
14:13:09 <jralls> As long as you don't expose the Win7 VM to the outside net much it's probably OK, and you might want to set up a new VM instead of upgrading the existing one. That way you can still test things on Win7 when necessary.
14:20:23 <gjanssens> That was the plan indeed.
14:21:10 <gjanssens> As for the bug, I don't see enough detail to point at something in particular (yet)
14:21:37 <gjanssens> If anything we may need profiling data from the reporter at some point.
14:25:31 <gjanssens> Hey, I just found how to configure kdevelop to properly deal with c++17 features \o/
14:25:51 <gjanssens> It was not a bug after all but a configuration I had overlooked.
14:26:09 <gjanssens> No more "namespace std doesn't define a variant class" errors :)
14:26:11 <jralls> Woo-hoo!
14:26:34 <gjanssens> Now we're cookin'
14:27:02 * jralls is looking forward to concepts, but that will have to wait for GC5.
14:27:57 <gjanssens> Back to the bug, I think it's up to the reporter again. I'll wait for his replies and will suggest a recipe for profiling with heaptrack on linux if needed.
14:28:08 <gjanssens> That's the profiler I'm most familiar with...
14:28:29 <gjanssens> Can even make diffs of two profile runs, which may be handy in this case
14:28:50 <jralls> OK. I use the MacOS Instruments one because it's really easy to use and doesn't require a special build.
14:29:20 <jralls> And I'm too lazy to learn a different one. I don't suppose heaptrack can be convinced to run under MinGW?
14:29:48 <gjanssens> No idea...
14:30:01 <gjanssens> But as Instruments it doesn't require a special build, which is good
15:09:05 *** frakturfreak has joined #gnucash
15:09:05 *** ChanServ sets mode: +v frakturfreak
15:16:20 *** Gerd has joined #gnucash
15:18:32 *** Gerd has quit IRC
15:19:12 *** Han has quit IRC
15:42:54 *** Mechtilde has quit IRC
15:45:17 *** fell has quit IRC
15:47:10 *** gjanssens has quit IRC
16:07:15 *** Gerd has joined #gnucash
16:14:04 *** Gerd has quit IRC
16:18:59 *** ramontjunior has quit IRC
16:56:00 *** lagertonne has quit IRC
17:02:12 *** sbluhm has quit IRC
17:06:46 *** paul has joined #gnucash
17:09:39 *** frakturfreak has quit IRC
17:10:39 *** Aussie_matt has joined #gnucash
17:13:06 <paul> Hello, I'm trying to import my customers.csv into gnucash 3.4-1+b10 on mx19. The notes column (just after the email column) is importing into a shipping column. The notes column is being skipped in the import gui. How do I change the csv so the notes column imports correctly?
17:14:50 <jralls> paul, I don't know and gjanssens has gone to bed. Ask on the gnucash-user mailing list.
17:16:29 <paul> Okay. Thanks. I'll look up on the website on how to do that. I see there's a 'mailing list' webpage so should be easy (never used a mailing list). Thanks again.
17:18:16 *** cal-tec has joined #gnucash
17:23:40 <paul> Mailing list exposes my email address to the interet. What's the next best support option that doesn't expose my email address to the internet?
17:24:02 *** cal-tec has quit IRC
17:26:53 * chris will drop conversation about QIF.
17:28:58 <jralls> paul, I don't know if you can set up a nabble account to do that, https://www.nabble.com/. If not you can create a burner email account someplace like gmail or hotmail.
17:30:49 <jralls> paul, the third option would be to lead off your question with "@tell gjanssens". That will save the message for him when he comes back. If you're not here he'll do the same and you'll get the message when you next create input here--several of us type a single '.' to retrieve messages.
17:37:55 <paul> Thanks. I'll likely do that. Or one of your other suggestions. First time using irc ... so if I want to retrieve messages for me I just type a '.'?
17:39:22 <chris> paul: or any other message
17:40:05 <chris> jralls: how's c++options going... looks like good progress to me, will be fine for 4.x?
17:40:26 <chris> potentially book-saved-reports for 5.x?
17:42:01 <jralls> I think so. I've just worked out how I'm going to do dialog-options, now it's just plug-and-chug for a couple of days... but I'm at a conference for a week starting tomorrow. I'll have it on my laptop but I don't know how much work I'll get done.
17:42:17 <paul> .
17:42:45 <chris> @tell paul test this
17:42:45 <gncbot> chris: The operation succeeded.
17:43:07 <jralls> paul, now type your dot.
17:43:12 <paul> .
17:43:12 <gncbot> paul: Sent just now: <chris> test this
17:43:13 <chris> enjoy conference!
17:43:52 <paul> thx
17:43:58 <chris> any thoughts what would happen to app-utils/*.scm
17:44:31 <jralls> Thanks. After that the next step will be to adjust the scheme API used in reports and then clean up. I hope to get it all done by the end of march.
17:46:11 * chris thinks fin.scm is a mess
17:51:30 *** chris1 has joined #gnucash
17:53:42 <paul> @tell gjanssens I'm trying to import my customers.csv into gnucash 3.4-1+b10 on mx19. The notes column (just after the email column) is importing into the shipname column. The notes column is being skipped in the import gui. How do I change the csv so the notes column imports correctly?
17:53:42 <gncbot> paul: The operation succeeded.
18:11:06 <jralls> chris is right.
18:11:38 * jralls is convinced that just about everything in libgnucash/app-utils is a mess.
18:11:43 <chris> lol
18:11:58 <chris> so... maybe fin.scm gets moved to bindings
18:12:05 <chris> date-utilities too
18:13:09 <jralls> Unfortunately fin.scm is like options: It's hybrid scheme+C that the register and loan assistant uses.
18:15:16 <jralls> Date-utilities OTOH is only used in report code IIRC so it could go to bindings or gnucash/reports.
18:41:37 <paul> @tell gjanssens Nevermind. I got it working. Had to make sure each line in the csv ended with a double semi-colon. Also, in the import gui, the notes column was still being skipped but once I clicked on the "semicolon separated with quotes" bullet and then clicked back on the "semicolon separated" bullet ... the notes column was not longer being skipped. Thanks to all the devs. Donated again this year. Cheers.
18:41:37 <gncbot> paul: The operation succeeded.
19:26:05 *** jralls has quit IRC
20:23:59 *** oozer has quit IRC
20:49:44 *** jervin has quit IRC
21:03:38 *** warlord2 is now known as warlord
21:16:50 *** paul has quit IRC
21:46:05 *** omnireq has quit IRC
21:51:46 *** chris1 has quit IRC
22:22:27 *** shaggy has left #gnucash
23:24:21 *** ericdm has joined #gnucash
23:25:41 *** ericdm has quit IRC
23:42:49 *** ericdm has joined #gnucash