2018-09-15 GnuCash IRC logs
00:00:45 *** greenshoe has joined #gnucash
00:21:16 *** nate has joined #gnucash
00:22:49 <nate> Hello World! Just was wondering how you guys managed your dev contributions. Do you only take patches, or are we allowed to commit to a dirty branch of any sort?
00:24:14 *** Mechtilde has joined #gnucash
00:57:32 *** greenshoe has quit IRC
01:16:35 *** gjanssens has joined #gnucash
01:16:35 *** ChanServ sets mode: +o gjanssens
01:16:38 <gjanssens> .
01:16:38 <gncbot> gjanssens: Sent 6 hours and 56 minutes ago: <jralls> Sorry, I didn't intend to be AFK all day, but had a plumbing leak that i needed to do a temp repair on.
01:16:39 <gncbot> gjanssens: Sent 6 hours and 54 minutes ago: <jralls> Interesting about xaccAccountRecomputeBalance. I'll have a look at the XML backend to see what it does to avoid that.
01:18:43 *** pilotauto has quit IRC
01:34:15 *** gncbot sets mode: +o fell
01:37:35 <fell> nate: Did you read https://wiki.gnucash.org/wiki/Development_Process?
01:40:38 *** fell_laptop has joined #gnucash
01:40:49 *** Aussie_matt has joined #gnucash
01:40:59 *** gncbot sets mode: +o fell_laptop
01:42:43 *** fell has quit IRC
01:42:50 *** fell_laptop is now known as fell
02:29:20 *** Mechtilde has quit IRC
02:39:08 *** fell has quit IRC
02:45:29 *** fell has joined #gnucash
02:48:03 *** Robert847 has left #gnucash
02:49:08 *** gncbot sets mode: +o fell
02:57:54 *** greenshoe has joined #gnucash
03:00:12 *** jotrago has quit IRC
03:00:54 *** greenshoe has quit IRC
03:38:34 *** jotrago has joined #gnucash
04:19:42 <fell> CMakeLists.txt does not mention boost_locale in its error text.
04:25:12 *** boldstripe has joined #gnucash
04:34:20 *** bertbob has quit IRC
04:37:13 *** bertbob has joined #gnucash
04:50:15 *** bertbob has quit IRC
04:55:37 *** chris has joined #gnucash
04:58:58 *** boldstripe has quit IRC
05:00:40 *** ncv has joined #gnucash
05:14:14 *** fabior has joined #gnucash
05:32:44 *** Mechtilde has joined #gnucash
05:46:12 *** storyjesse has quit IRC
06:06:05 *** fekepp has joined #gnucash
06:07:24 *** ncv has quit IRC
06:13:06 *** fekepp has quit IRC
06:15:41 *** ncv has joined #gnucash
06:20:18 *** ncv has quit IRC
06:20:34 *** ncv has joined #gnucash
06:23:34 *** ncv has quit IRC
06:24:26 <fell> 8 month after dropping autotools, I got my first CMake build in Eclipse running. :-)
06:30:13 *** O01eg has joined #gnucash
06:32:07 *** oozer has joined #gnucash
06:38:55 *** ncv has joined #gnucash
06:39:21 *** O01eg has quit IRC
06:40:06 *** ncv has quit IRC
06:47:24 *** jotrago1 has joined #gnucash
06:47:34 *** jotrago has quit IRC
06:47:34 *** jotrago1 is now known as jotrago
06:50:44 *** fekepp has joined #gnucash
06:53:04 *** O01eg has joined #gnucash
06:55:54 *** greenshoe has joined #gnucash
06:56:21 *** greenshoe has joined #gnucash
06:59:21 *** greenshoe has quit IRC
07:03:54 *** fabior has quit IRC
07:09:18 <gjanssens> fell: congrats :)
07:23:31 <fell> Until the 2.6.21 release there was still enough to do on maint, so I never tried before. And in between I got too often interrupted by discussions...
07:26:13 *** jotrago1 has joined #gnucash
07:26:34 *** jotrago has quit IRC
07:26:34 *** jotrago1 is now known as jotrago
07:29:40 *** fabior has joined #gnucash
07:44:08 *** jotrago1 has joined #gnucash
07:44:34 *** jotrago has quit IRC
07:44:34 *** jotrago1 is now known as jotrago
07:45:06 *** greenshoe has joined #gnucash
07:54:35 *** Mechtilde has quit IRC
08:12:07 *** Mechtilde has joined #gnucash
08:26:38 *** oozer has quit IRC
08:27:03 *** Jimraehl1 has joined #gnucash
08:27:40 *** Jimraehl1 has left #gnucash
09:09:59 <fell> chf: Du hast Post.
09:11:38 <fell> and can you add linebreaks in the namespace section?
09:36:16 *** ChattyMan has joined #gnucash
09:51:30 *** jethrob has quit IRC
10:02:55 *** ChattyMan has quit IRC
10:10:25 *** jethrob has joined #gnucash
10:19:19 *** greenshoe has quit IRC
10:20:43 *** greenshoe has joined #gnucash
10:23:43 *** greenshoe has quit IRC
10:28:15 *** fabior has quit IRC
10:28:35 *** fabior has joined #gnucash
10:31:45 *** fabior has quit IRC
10:34:07 * fell is wondering about "<slot xmlns=""><slot:key>color</slot:key><slot:value type="string">Not Set</slot:value></slot></act:slots>"
10:35:12 <fell> 1. Was it intended that way?
10:35:36 <fell> 2. Do we want it in account templates?
10:37:22 *** greenshoe has joined #gnucash
10:44:43 <gjanssens> fell: that looks like superfluous. Though I'm how the code would react if the slot is completely omitted. Can you test ?
10:45:14 <gjanssens> jralls: https://github.com/gjanssens/gnucash/commit/544341690573e992065749b4ed57e176c33927df seems to do the trick wrt the xaccAccountRecomputeBalanceInCurrency trashing
10:46:05 <gjanssens> It drops the load time from 53m to 1m20s for Monster DB™
10:46:27 <gjanssens> I haven't pushed to maint yet though as I would like your review first
10:47:02 *** fabior has joined #gnucash
10:47:11 <gjanssens> I'm hesitating due to introducing old glib style code in the sql backend (a glist :(
10:47:26 <gjanssens> I didn't immediately see another way though.
10:47:49 <gjanssens> Anyway since you're more at home in the sql backend, I prefer you look at it first.
10:51:39 <gjanssens> Obviously the spped improvements will depend on many factors.
10:52:21 <gjanssens> I'm on a fast machine with a fast ssd. On average machines with normal HDD's the improvement will not be equally dramatic
10:54:18 *** fabior has quit IRC
10:58:12 *** greenshoe has quit IRC
10:58:46 *** Aussie_matt has quit IRC
10:59:23 *** Mechtilde has quit IRC
11:00:30 *** oozer has joined #gnucash
11:01:19 *** Mechtilde has joined #gnucash
11:05:03 <gjanssens> Oh I just found gnc_account_for_each_descendant, which actually does what I want. I'll use that instead.
11:06:50 *** bertbob has joined #gnucash
11:08:09 *** jotrago1 has joined #gnucash
11:08:34 *** jotrago has quit IRC
11:08:34 *** jotrago1 is now known as jotrago
11:10:08 <gjanssens> New implementation: https://github.com/gjanssens/gnucash/commit/5775662b52869604913a44972737cd8748945169
11:10:33 <gjanssens> I suspect the choice of loading transactions before business objects is also important here.
11:11:16 *** fabior has joined #gnucash
11:12:59 <gjanssens> Before I did that the remainder loading did happen with raised account edit levels according to https://github.com/gjanssens/gnucash/blob/maint/libgnucash/backend/sql/gnc-sql-backend.cpp#L297
11:13:55 <gjanssens> Unless the business loading functions already forced the load of many splits of course, in which case the bashing would happen while loading business objects.
11:14:20 *** fabior has quit IRC
11:16:06 *** fabior has joined #gnucash
11:18:44 *** greenshoe has joined #gnucash
11:26:55 *** Mechtilde has quit IRC
11:28:11 *** Mechtilde has joined #gnucash
11:35:53 *** Mechtilde has quit IRC
11:37:42 *** Mechtilde has joined #gnucash
11:41:03 *** Mechtilde has quit IRC
12:47:12 *** boldstripe has joined #gnucash
12:55:35 <fell> In older templates, I see i.e. <act:slots><slot><slot:key>placeholder</slot:key><slot:value type="string">false</slot:value></slot> ... </act:slots>, but the version which chf used to update SKR49 adds the empty namespace attribute. (Still researching...)
12:59:07 <gjanssens> jralls: perhaps we should extend the account edit level manipulation to include the business object loading also
13:03:47 *** chf has quit IRC
13:04:18 *** chf has joined #gnucash
13:08:44 *** greenshoe has quit IRC
13:14:50 <gjanssens> Anyway, no need for a review I guess, I merely redid what was apparently already done in another part of the code
13:15:32 *** greenshoe has joined #gnucash
13:16:11 *** btr has joined #gnucash
13:20:42 <btr> I wonder if I found a bug: Was using GnuCash 2.6.21 on WinXP w/ SQLLite. Switched to GnuCash 3.2 on Win10 w/ SQLLite and I am unable to update scheduled transactions. The stated error is that the transaction is already being updated in another window. Converting back to the XML file format seems to resolve the issue. No other observed issues in either configuration.
13:30:17 <jralls> gjanssens: I'm more comfortable with the second version. I was thinking earlier about getting a list of accounts to wrap with "SELECT DISTINCT account_guid FROM SPLITS" but just doing the whole tree is less gyrating.
13:31:55 *** Mechtilde has joined #gnucash
13:32:32 <jralls> gjanssens: As for the business objects, it shouldn't be necessary. The point of forcing the transaction load as early as possible was to shut down the one-transaction-at-a-time loading.
13:34:40 <jralls> Once all of the transactions and their splits are in place then the biz objects just retrieve the ptrs from the QofID containers.
13:37:10 <jralls> btr: It's https://bugs.gnucash.org/show_bug.cgi?id=796248, fixed for 3.3 that will be released in ~2 weeks.
13:37:48 <btr> jralls: Thanks!
13:38:33 <jralls> btr: Since you're on Windows you can grab any recent nightly from https://code.gnucash.org/builds/win32/maint
13:42:00 <btr> jrails: Thanks. I just might do that if I can scrounge the spare time (need to go look up what that is)
13:42:40 <KevinDB> are there mac nightlies?
13:43:23 <jralls> KevinDB: No, warlord doesn't have a mac to do them on.
13:44:02 <KevinDB> In that case I shall wait patiently :-)
13:45:54 *** fekepp has quit IRC
13:46:16 *** fekepp has joined #gnucash
13:50:28 *** fabior has quit IRC
13:54:39 *** gncbot has joined #gnucash
13:57:10 *** chf has quit IRC
13:57:31 *** chf has joined #gnucash
13:57:42 *** CDB-Man has quit IRC
14:08:58 *** boldstripe has quit IRC
14:10:21 *** greenshoe has quit IRC
14:12:49 *** greenshoe has joined #gnucash
14:13:46 *** greenshoe has joined #gnucash
14:14:02 *** btr has quit IRC
14:14:35 <fell> Found it! It is a sideffect of using xsltproc gnc2xea.xsl ...
14:25:51 *** frakturfreak has joined #gnucash
14:30:34 *** jotrago has quit IRC
14:30:39 *** jotrago has joined #gnucash
14:33:10 *** jotrago has quit IRC
14:33:13 *** jotrago has joined #gnucash
14:39:50 *** greenshoe has quit IRC
14:45:02 *** greenshoe has joined #gnucash
14:46:42 *** greenshoe has quit IRC
14:47:02 *** greenshoe has joined #gnucash
14:48:14 *** jralls sets mode: +o gncbot
14:50:02 *** greenshoe has quit IRC
15:00:30 *** greenshoe has joined #gnucash
15:00:46 *** fekepp has quit IRC
15:00:47 *** fekepp has joined #gnucash
15:18:34 *** boldstripe has joined #gnucash
15:19:28 *** boldstripe has quit IRC
15:20:22 *** greenshoe has quit IRC
15:20:45 *** greenshoe has joined #gnucash
15:27:23 *** fekepp has quit IRC
15:27:24 *** fekepp has joined #gnucash
15:37:03 <chris> I have idea about refactoring while keeping API safe for custom report users -- just copypasta the API into the calling report, and deprecate the API with warning "gnc:some-function is deprecated and will be removed before 2020." but it would be nice to display this warning prominently instead of Tracefile.
15:37:14 *** greenshoe has quit IRC
15:47:24 <chris> ^ this can be used for the html-chart upgrade, keeping old html-bar|line|scatter|pie around for 1 major release cycle
15:47:37 <jralls> chris: I don't understand how moving a deprecated function from report-system to standard-reports won't break a user's custom report.
15:48:31 <chris> well it won't if i keep the report-system function intact and paste the refactored one in the standard-report
15:51:06 <chris> I'll try on master-chartjs.... keeping old api alive while promoting new api. i think old api can even call new api.
15:52:14 *** fell has quit IRC
15:52:18 <jralls> A *refactored* function has the same signature, it has guts that are "less smelly" (to borrow from Fowler). If you're replacing API it's not really refactoring.
15:53:21 *** Mechtilde has quit IRC
15:53:22 <jralls> When you replace API you want to mark the old API deprecated. Client code then has time to adopt the new API before the deprecated functions are removed.
15:58:51 <chris> hmm will need to thnk about old charts api.
15:59:08 <chris> they can probably directly call new html-chart api and display a decent chart.
16:00:41 <chris> because my #316 is now known to be drastic, cutting off old api which can be remapped onto new one
16:01:35 <chris> (define gnc:html-barchart-set-title! gnc:html-chart-set-title!) should do the trick
16:03:58 <chris> ps we have our japan tickets 14-24 nov
16:05:33 <jralls> I just got our tickets last week, 4 Nov-5 Dec.
16:06:03 <chris> cool, may our paths cross
16:07:02 <jralls> We'll be in Kyoto the whole time you're there, should be possible to meet up if it's on your itenerary.
16:07:54 <chris> oh definitely... Kyoto appeals to the wife, not sure why
16:08:04 <jralls> BTW, do you know about JR (Japan Railway) passes? You have to get them before you leave. They're a huge savings if you're going to be touring around.
16:08:38 <chris> yeah i know, visited back in 2005
16:08:52 * chris needs back to sleep
16:09:08 <jralls> OK, goodnight.
16:32:01 *** gjanssens has quit IRC
16:45:55 *** frakturfreak has quit IRC
16:52:12 *** CDB-Man has joined #gnucash
16:53:17 *** CDB-Man has quit IRC
16:55:35 *** CDB-Man has joined #gnucash
17:01:41 *** frakturfreak has joined #gnucash
18:00:32 *** fekepp1 has joined #gnucash
18:02:23 *** fekepp has quit IRC
18:02:23 *** fekepp1 is now known as fekepp
18:06:24 *** chf has quit IRC
18:07:02 *** chf has joined #gnucash
18:21:37 *** frakturfreak has quit IRC
18:22:55 *** frakturfreak has joined #gnucash
18:40:43 *** JayC has quit IRC
20:09:55 *** fekepp has quit IRC
20:21:34 *** nate has quit IRC
20:23:01 *** nate has joined #gnucash
20:39:22 *** fiddlerwoaroof has quit IRC
20:53:51 *** fiddlerwoaroof has joined #gnucash
21:01:33 *** oozer has quit IRC
21:44:09 *** frakturfreak has quit IRC
23:14:45 * chris thinks - the #392 balsheet-pnl perhaps doesn't need to be merged - i'll try transform existing code into it
23:15:43 * chris also #316 will be transformed to be less drastic - will maintain existing charts api which will map onto html-chart one
23:22:31 *** fell has joined #gnucash
23:30:32 *** fell has quit IRC
23:31:14 *** fell has joined #gnucash
23:31:14 *** gncbot sets mode: +o fell