2019-02-02 GnuCash IRC logs
00:30:25 *** chris has quit IRC
00:30:29 *** chris has joined #gnucash
00:31:27 *** Aussie_matt has joined #gnucash
00:55:48 *** chris has joined #gnucash
01:33:29 *** fell_laptop has joined #gnucash
01:35:12 *** fell has quit IRC
01:46:59 *** fell_laptop has quit IRC
02:01:42 *** gjanssens has joined #gnucash
02:01:42 *** ChanServ sets mode: +o gjanssens
02:02:29 *** fell has joined #gnucash
02:04:38 *** fabior has joined #gnucash
02:44:39 *** fabior has quit IRC
02:57:20 *** kapil___ has joined #gnucash
03:20:06 *** bertbob has quit IRC
03:22:29 *** bertbob has joined #gnucash
03:38:41 *** boldstripe has joined #gnucash
03:40:33 *** matt has joined #gnucash
03:40:35 *** gour has joined #gnucash
03:55:52 *** boldstripe has quit IRC
03:58:24 *** boldstripe has joined #gnucash
04:04:24 *** matt has quit IRC
04:32:11 <gjanssens> .
05:32:03 *** Aussie_matt has quit IRC
05:40:12 *** User__ has joined #gnucash
05:54:06 *** boldstripe_ has joined #gnucash
05:54:20 *** boldstripe has quit IRC
05:54:20 *** boldstripe_ is now known as boldstripe
05:56:26 *** kapil___ has quit IRC
05:56:55 *** fabior has joined #gnucash
05:59:03 *** monkeyjuice has joined #gnucash
06:00:36 *** monkeyjuice has quit IRC
06:01:31 *** Aussie_matt has joined #gnucash
06:12:30 *** boldstripe has quit IRC
06:31:52 *** finster has joined #gnucash
06:39:46 *** User__ has quit IRC
06:43:39 *** fabior has quit IRC
06:44:51 <finster> Hello all. Not an accounting pro, so this may be a dumb question: I'm using a standard report showing net worth for the current accounting period. It shows the net worth per e.g. 1st of February, But all transactions of February 1st are included in the net worth value. I would think the net worth should be the one ending January 31st, not February 1st, if that makes any sense
06:45:50 <finster> Say i have a net worth of USD 10000 at the end of January 31st. Then spend USD 1000 on February 1st, the report would show a net worth of USD 9000 for Feb 1st, which is confusing at least
06:46:55 <finster> While technically correct, I suppose
06:47:40 <finster> Thanks for releasing Gnucash 3.4 btw!
06:47:48 <chris> finster: are you https://bugs.gnucash.org/show_bug.cgi?id=797076 ?
06:49:06 <finster> chris: no
06:50:01 <chris> ok valid question, easy fix, but no easy fix that will be 100% backward compatible with current behaviour
06:51:27 <finster> Okay, so my logic was not entirely off :
06:51:48 <chris> the 'sensible' fix would be to choose start-date of 31-january but this causes lots of problems when reaching february
06:52:04 *** Guest28 has joined #gnucash
06:52:38 *** Guest28 is now known as Couto
06:53:38 <chris> so, the assumption of 'net-worth on date X includes transactions on date X' seems pervasive from the design of reports
06:53:52 <chris> I have no good answer yet
06:54:19 <finster> Alright, that's okay
06:55:36 *** monkeyjuice has joined #gnucash
06:56:51 *** monkeyjuice has quit IRC
07:00:39 <Couto> I bet a lot of people ask this, but im a total noob with finances and im trying to get a grip on all this... Does anyone recommend a good resource on how to start with gnucash when we already have a set of bank accounts and loans?
07:03:51 <finster> Couto: from my experience about 2 years ago (also a finance noob), be prepared to change your account setup a lot in the first weeks
07:04:25 <chris> Couto: the Tutorial and Concepts Guide is incredibly well written
07:05:50 <Couto> chris I've taken a look at the tutorial and concept guides (and I really have to praise them, i've rarely seem a software which such great documentation) However, I do have to admit that my lack of financial knowledge is getting in the way :( I seem to get if I was starting from zero, but imagine (e.g) a house loan that's already half-paid
07:05:58 <Couto> that's where Im getting stuck
07:06:17 <Couto> finster yeah, I've tried twice already and i already changed a couple of things
07:07:15 <chris> Well my approach would be a single transaction with 3 splits: e.g. Asset:House $100,000, Liability:Mortgage -$35,000, Equity:Opening -$65,000
07:08:10 <chris> or you could do 2 txns: equity->asset $100,000; equity<-liability $35,000, both dated on opening balance day
07:12:57 <Couto> chris out of curiosity, is not possible to insert past transactions right? Imagine that I create my gnucash file right now, and insert my current bank balance (as in today).... it doesnt make to sense to create a home loan before today's date and insert all payments and transactions relative to that loan right?
07:13:47 <chris> you can if you want, you'll need to adjust your "opening balance" transaction
07:16:02 <Couto> makes sense :|
07:16:33 <Couto> maybe your first suggestion of a split transaction makes more sense
07:21:32 *** couto[m] has joined #gnucash
07:22:18 *** couto[m] is now known as Couto[m]
07:23:07 *** Couto has quit IRC
07:23:40 <Couto[m]> I really wouldnt mind to pay for a video course about basic home finance and gnucash :|
07:23:44 *** puck has quit IRC
07:26:24 *** puck has joined #gnucash
07:59:35 <chris> Couto: if you wish to 'insert past transactions', one good trick is to free-text the opening balance amount in the description. e.g. in the house opening bal above, Liability:Mortgage -$35000 Equity:Opening $35000, description "Opening Bal is $35000 on 01.01.2018"
08:00:07 <chris> because if in the future you wish to add the past transactions, the 01.01.2018 amount won't make sense anymore
08:06:31 *** nimish has joined #gnucash
08:33:03 *** oozer has joined #gnucash
08:33:41 *** Jimraehl1 has joined #gnucash
08:33:50 *** trying2Simplify has joined #gnucash
08:34:14 *** Jimraehl1 has left #gnucash
08:37:50 <gjanssens> finster: Why would a net worth go back a day ?
08:38:13 <gjanssens> The net worth of the current period is up until this very moment
08:38:38 <gjanssens> If you want to know the net worth on January 31st, you'd ask for the net worth on January 31st.
08:39:13 <gjanssens> The day you select is included, that is, the code assumes "end of day"
08:39:33 <gjanssens> chris I quickly looked at bug 797076
08:39:45 <gjanssens> But I'm not sure I see the issue
08:39:54 <chris> can't see either
08:40:14 <gjanssens> At least I don't see a problem in the screenshots presented
08:40:47 <gjanssens> All dates seem to match as far as I can tell, unless perhaps the very last two bars, which may be duplicate
08:40:58 *** Aussie_matt has quit IRC
08:41:18 <gjanssens> If the issue is with actual calculated values, there are no details on the report to go by.
08:41:38 <gjanssens> Let me ask for a small sample file illustrating the issue...
08:41:54 <finster> gjanssens: i'm looking at net worth in retrospective, say for the last year. the report states the net worth for the 1st of every month, insinuating the value shown is right at the end of each month, which is it effecitvely not
08:42:41 <chris> finster's issue is valid for networth-linechart eg end-of-month networths ... 31/10 -> 31/11 -> 31/12 -> 31/01 -> ?? 2/3 ???
08:43:31 <chris> oops 31/10 -> 01/12 -> 01/01
08:43:58 <gjanssens> chris 31/10 -> 30/11 -> 01/12 -> 01/01 then ?
08:44:32 <chris> well exactly 31/10 -> 01/12 because there's no 31/11
08:44:33 <gjanssens> Anyway ISTR the jump in february/march has come up in the past. It may be a known bug or a regression
08:44:51 <gjanssens> Oh, right
08:44:52 <chris> yes ive seen before
08:45:07 <gjanssens> Should still be fixed though
08:45:11 <gjanssens> :)
08:46:06 <chris> ^ the only sane method is to think of a 'end-of-month' interval specifier so it chooses 31/10, 30/11, 31/12, 31/01, 28(or 29)/2, 31/3 etc
08:46:22 <chris> but, not easy ;)
08:47:40 <gjanssens> as far as I know that exists
08:47:50 <gjanssens> The issue is mostly in how the next interval is calculated
08:48:04 <gjanssens> I think that code should take the length of the current month into account
08:48:28 <gjanssens> IIRC the issue came up before in the SX region of gnucash
08:48:39 <finster> As for my issue, it's not only a problem for the current month, but for the past months, too
08:48:39 *** jervin has joined #gnucash
08:49:51 <chris> well i cantry fix reports but not SX ;)
08:49:53 *** jervin has quit IRC
08:51:01 *** g5pw has quit IRC
08:51:24 <gjanssens> Actually I saw it before in the calculation of the next reconcile period: https://bugs.gnucash.org/show_bug.cgi?id=737099
08:51:50 *** g5pw has joined #gnucash
08:51:58 <gjanssens> finster: can you share a screenshot ?
08:52:25 <finster> I'll prepare a new book. Allow for some minutes
08:56:35 <chris> meanwhile here's screenshot for next-gen charts <a href> anchor: https://imgur.com/a/Mg3twu2
08:59:03 <finster> https://imgur.com/gallery/WZ9vUCo
09:01:26 <finster> So, the screenshot shows the dates correctly (e. g. Aug 31st), when a transaction as made at September 1st.
09:01:38 <finster> Correct so far (imo)
09:02:36 <chris> would you mind posting the barchart version ?
09:04:33 <finster> Bingo!
09:05:24 <finster> that does in deed show "only" the 1st of each month: https://imgur.com/gallery/iXCFmcX
09:06:34 <chris> Here's a relevant bugs
09:06:35 <chris> https://bugs.gnucash.org/show_bug.cgi?id=796895
09:06:49 <chris> there's a patch to html-linechart.scm which you can try
09:07:55 <chris> as well as the explanation
09:08:09 <finster> Alright, i'll give it a go
09:09:56 <finster> Thanks for spending the time so far!
09:10:43 <chris> np
09:15:38 *** boldstripe has joined #gnucash
09:19:24 <gjanssens> Perhaps I'm slow, but why would you think the line chart is showing correct dates ?
09:19:40 <gjanssens> To me it looks like it's jumping around a lot
09:20:15 <gjanssens> Wouldn't it make sense to have all dates consistently be the first or last date of the month ?
09:21:58 <chris> gjanssens - this is not finster's fault
09:22:19 <chris> html-chart is receiving data for linechart via x=date-string, y=number
09:23:07 <gjanssens> I wasn't blaming finster :)
09:23:18 <chris> jqplot seems to try to divide the whole duration (startdate->enddate) into N equal blocks, then chooses dates to fit each block boundary
09:23:20 <gjanssens> I was trying to figure out what is to be expected
09:23:40 <gjanssens> Right. jqplot limitations...
09:23:47 <gjanssens> How is this in chartjs ?
09:23:51 <chris> no: not jqplot limitatoin
09:23:57 <chris> html-linechart was written as such
09:24:10 <gjanssens> Sorry, knee jerk reaction
09:24:38 <gjanssens> So linechart's period calcuation is flawed as well here?
09:24:42 <chris> no
09:25:55 <gjanssens> No?
09:25:55 <chris> it's sending data appropriately, 30/10, 30/11, 30/12 ...etc to jqplot which interprets these datestrings approrpiately as x-axis coordinate
09:26:47 <chris> so if we force jqplot to stop interpreting datestrings as 'coordinate' and force 'categories' instead then problem solved
09:27:28 <gjanssens> Ok I'm with you so far
09:27:38 <chris> so even if html-linechart is sending regular 01-jan 01-feb 01-mar to jqplot,
09:27:53 <chris> ^(scratch that)
09:28:37 <chris> the current behavior sends data as datestring=date=x coordinate
09:28:51 *** fabior has joined #gnucash
09:29:10 <gjanssens> So the point really is we shouldn't be using x coordinate style x-axes, but always category style x-axes
09:29:16 <chris> bingo
09:29:31 <gjanssens> With coordinate style, we'll always run into rounding confusion
09:29:37 <chris> bingo
09:29:45 <gjanssens> And presumably that would still be so with chartjs
09:29:49 <chris> bingo
09:30:07 <gjanssens> 3 bingos in a row... what did I win ? ;)
09:30:25 <chris> with jqplot you'll then run the error of 'crowded xAxis labels' which you fixed a long time ago via an ugly hack
09:30:36 * gjanssens remembers that one
09:30:40 <chris> and chartjs is slightly more clever :)
09:30:46 <gjanssens> yay ?
09:30:50 <gjanssens> eh yay!
09:31:49 <gjanssens> For a moment I was confused as finster was talking about 1st of each month and you came up with 30/10, 30/11, 30/12...
09:32:08 <gjanssens> Your example was not related to finster's ?
09:32:18 <chris> i was initially confused too
09:33:01 <gjanssens> Still as he mentioned initally: he was expecting 31/1 but saw 1/2
09:33:08 <gjanssens> Is that then a separate issue ?
09:33:34 <gjanssens> Should we be presenting net worth at end-of-period dates instead of start-of-period ?
09:34:07 <gjanssens> Because net worth is calculated at end of day of the date specified
09:34:16 <chris> this is a matter of preference... your networth today should include today's earnings?
09:34:36 <gjanssens> That's only a question for today
09:35:08 <gjanssens> If you speak of net worth at last book closing that would be net worth on Dec 31st 23:59
09:35:19 <gjanssens> So the last day of that period
09:35:23 <gjanssens> Not the first day of the next
09:35:34 <chris> but the crux is, if someone wishes to display monthly networth for 01-jan 01-feb 01-mar, they're really expecting to exclude txns on the 1st of each month, which I think is a fair comment
09:35:47 <chris> so our hack to ask them to choose 31-dec as starting point won't ever work
09:36:11 <gjanssens> Why nt ?
09:36:14 <gjanssens> Why not?
09:36:21 <chris> 31-dec, 31-jan, 31-FEB?!?
09:36:24 *** trying2Simplify has quit IRC
09:36:30 <chris> this old chestnut
09:36:47 <gjanssens> That's gnucash' problem to handle not the users
09:37:09 <gjanssens> It should be clever enough to know how many days are in each given month
09:37:10 <chris> yes and it's still an unfixed quirk
09:37:15 <gjanssens> Indeed
09:38:08 <gjanssens> Going the other way around as you suggest is IMO not a good idea (that is ask on 1st and report on 31st)
09:39:06 <gjanssens> Almost everywhere in gnucash where we use intervals this is from start-of-day on the first day to end-of-day on the last day
09:39:41 <gjanssens> It would require huge amounts of date time rewriting to change this
09:40:01 <chris> ... and writing suitable tests
09:40:55 <gjanssens> So to be clear, in gnucash internally a monthly interval currently means from Jan 1st 0:00 to Jan 31st 23:59, Feb 1st 0:00 to Feb 28/29th 23:59 and so on
09:41:18 <gjanssens> It's not Jan 1st 0:00 to but excluding Feb 1st 0:00
09:41:54 <chris> afaiu in reports this is enforced pretty much all of them
09:42:12 <gjanssens> Which is enforced ?
09:43:01 <chris> (gnc:start-day-time start-date) and (gnc:end-day-time end-date)
09:44:45 <gjanssens> Ok, that's what I meant
09:45:15 <gjanssens> And what would require a lot of work and care to change
09:45:20 <chris> outside reports i have no idea!
09:47:17 <gjanssens> However - and now it gets confusing - net worth is not an interval calculation, it's calculated at specific points in time
09:47:30 <gjanssens> Not over a given period
09:48:03 <gjanssens> So what point in time should be considered when a user asks for the net worth on a given date
09:48:24 <chris> currently net-worth point is chosen to be end-of-date
09:48:42 <gjanssens> end-of-date ?
09:49:03 <chris> yes, networth on 01-jan means 01-jan 23:59
09:49:05 <gjanssens> I know end-of-period-date or end-of-day-time
09:49:28 <chris> brb 5min
09:49:31 <gjanssens> Ok so that's what I would expect, though 01-jan is an odd choice
09:50:21 <gjanssens> See, we define accounting periods as meaningful time periods
09:51:07 <gjanssens> I'd be interested to know net worth at the end of that period, so the shortest mental jump for me would be ask for net worth on that last day
09:51:40 <gjanssens> And that's how it is currently implemented as well
09:53:47 <chris> for balance-sheet it's not an issue
09:53:54 <chris> we can choose date anytime
09:54:06 <chris> it's the networth barchart that'll cause issues
09:54:40 <gjanssens> How is that different ?
09:55:03 <chris> because i may want monthly networth beginning of month
09:55:26 <gjanssens> Yeah I was about to make a note of that as well
09:55:59 <gjanssens> It looks like the choice of date selectors doesn't match a net worth bar chart very well
09:56:21 <chris> yes
09:56:31 * chris wish i was in brussels rn
09:56:44 <gjanssens> ?
09:56:50 <chris> to hear the great wingo talk
09:56:57 <gjanssens> Ah at fosdem
09:57:10 <gjanssens> You can follow it on live stream I believe :)
09:57:25 <chris> exactly right now
09:57:31 <chris> https://live.fosdem.org/watch/k4201
09:57:37 <gjanssens> I'll leave you to that then ?
09:58:02 <chris> it's fine can't hear much.
09:59:24 <gjanssens> I can, but of course Brussels is only 15km from here ;)
09:59:42 <gjanssens> Unfortunately this year I won't have time to attend fosdem.
10:00:31 <chris> time always an issue unfortunately.
10:00:38 <gjanssens> So on to the net worth charts, what would make a good date selector ?
10:01:18 <chris> hmm i'll need to think about that offline a bit, especially i don't want to make a net-charts.scm specific solution that won't apply elsewhere.
10:01:21 <gjanssens> Instead of "Start of previous month", should it simply be "Previous month" ?
10:02:08 <gjanssens> I think the key issue is we're trying to use interval oriented date selectors for point-in-time date selection
10:02:32 <gjanssens> "Start of" and "End of" don't make sense for point in time data
10:03:10 <chris> it's fine because we need an interval for the net-worth *chart*
10:03:45 <chris> i suspect some new-option is in order but i'll try and find a backward-compatible fix
10:04:16 <gjanssens> No the interval is the time difference between the points in time we want to chart
10:05:20 <gjanssens> And that is defined by the interval size in a separate option already (Step size)
10:05:50 <chris> yes... and this is the one that may have to change
10:06:13 <gjanssens> In what sense ?
10:07:16 <chris> currently specifies interval such as monthly/weekly, and we may need to make a special case for monthly... or we may need to modify gnc:make-date-list to produce better dates
10:08:27 <chris> i'll need to think about that a bit, because it'll also effect my unmerged multicolumn balsheet-pnl
10:10:09 <finster> Just returned from lunch. Wow, didn't expect the topic to blow up like this.
10:10:32 <chris> lol this is why modifying/upgrading anything in gnucash is not easy
10:10:44 <chris> 20yrs of baggage
10:11:45 <finster> Naive approache for monthly net worth stats: use the 1st of each month, go back 1 day, and display the net worth for that day?
10:14:22 *** monkeyjuice has joined #gnucash
10:16:20 *** monkeyjuice has quit IRC
10:21:34 *** kapil___ has joined #gnucash
10:21:39 *** Robert847 has left #gnucash
10:23:19 <chris> finster the main issue is backward compatibility
10:28:32 *** boldstripe has quit IRC
10:31:54 *** boldstripe has joined #gnucash
10:32:07 *** Mechtilde has joined #gnucash
10:34:38 <gjanssens> finster: that's something you can do now already if you want
10:35:02 <gjanssens> (minus the end of month bug on february)
10:35:30 <gjanssens> chris: I still don't understand why we need a special case for monthly ?
10:35:55 <chris> the issue is gnc:make-date-list
10:36:39 <chris> ... which calls (incdate date FooDelta)
10:37:20 <chris> MonthDelta causes (month + 1)
10:38:15 <chris> which is can cause 31-feb which is unacceptable to gnc-mkterm
10:39:11 <chris> so if we handle monthly differently (optional, via some flag)
10:39:39 <chris> it can send acceptable end-of-month dates to gnc_mkterm
10:40:01 <chris> so, selecting start-date 31-dec interval monthly becomes sane
10:41:15 <chris> and also month has to be a special case, because
10:41:54 <chris> WeekDelta 23-feb + 7days = 2-mar is absolutely acceptable
10:43:39 <gjanssens> Yes, agreed
10:43:53 <gjanssens> Though why would you need an option for this ?
10:44:13 <chris> hmm good point
10:45:06 <gjanssens> Though maybe you do
10:45:44 <gjanssens> as the function is called recursively and won't know about the initial date
10:46:28 <chris> perhaps the special case is not about 'monthly'
10:46:34 <gjanssens> Which you'd need to figure out what day to set for the month after February
10:47:01 <chris> it'll be 'any FooDelta which modifies month'
10:47:02 <gjanssens> ISTR we discussed this similar issue in the past with jralls and fell
10:47:12 <gjanssens> Yes, indeed
10:47:42 <chris> So, MonthDelta, QuarterDelta, HalfYearDelta
10:47:51 <chris> ^ bingo
10:48:36 <chris> I may wish to tack on the end of #450 if I can fix this tomorrow
10:48:38 <gjanssens> The difficult part is how to interpret dates with a day >= 28
10:48:50 *** boldstripe has quit IRC
10:49:20 <gjanssens> For example when a user asks for net worth on Feb 28 and following did he mean end of month or 28th of any month ?
10:49:46 <gjanssens> Same if s/he asked for April 30, was that end of month or 30th of every month ?
10:50:10 <chris> I think the sane approach is to do 28 of every month, and leave this as a limitation of choosing 28/feb as start-date
10:50:56 <gjanssens> Really the sane approach is to ask the user...
10:51:06 <gjanssens> But our code is not set up to do so at that point
10:51:24 <gjanssens> And it would mean a lot of overhead to implement that possibility
10:52:18 <gjanssens> Anyway I'm fine with either option as we don't have a way to ask the user
10:53:08 <gjanssens> Though thinking more of it I'd do it the other way around actually :)
10:53:47 <gjanssens> There are 7 months with 31 days and 5 months with less than 31.
10:53:50 <chris> a very naive but effective approach is: (1) add 1 to month, (2) call gnc_mkterm with new date, (3) if resulting date's month is not old-month+1 then minus 1 day from date, and goto (2)
10:54:10 <chris> the other approach is via a lookup table
10:54:49 <gjanssens> Selecting last day for each of those with not 31 days would result in all 31-day months not being treated as last day.
10:55:20 <gjanssens> And I actually think most people would want to see net worth on last-day rather than an arbitrary day
10:56:05 <chris> ouch the naive approach will fail because it'll go 31-dec 31-jan 28-feb 28-mar 28-apr
10:56:20 <gjanssens> you already tried ?!
10:56:27 <chris> in the head...
10:57:02 <gjanssens> Yes, but that will be an issue with a lookup table as well and is an inherent flaw in the current recursive approach of gnc:make-date-list
10:57:22 *** Mechtilde has quit IRC
10:57:28 * chris will attack gnc:make-date-list tomorrow >:)
10:57:49 <gjanssens> Once you pass a month with fewer days than the one you started with, the next month will no longer know it should be last day
10:58:59 <chris> ok i'll try an implementation tomorrow, to be tacked on at end of PR #450, and I think should handle month deltas differently
10:59:09 <gjanssens> I think there is your optional parameter to set on the very first call to gnc:make-date-list: if the initial date turns out to be last day of that month
10:59:32 <gjanssens> set a last-day flag to #t, otherwise set it to #f
10:59:43 <chris> in net-charts.scm?
10:59:45 <gjanssens> if it's set, don't reset it in subsequent calls
11:00:32 <gjanssens> I'd make it an optional parameter to gnc:make-date-list itself and only use set it from within that function if not set
11:00:40 <gjanssens> (if such a thing is possible in guile)
11:00:59 <gjanssens> That is you'd need to be able to distinguish between unset and #f
11:01:00 * chris cracks fingers
11:01:07 <gjanssens> :D
11:01:23 * gjanssens too, as they hurt from all this typing...
11:01:35 <chris> yes via (define* (gnc:make-date-list start end incr #:key (last-date? #f)) ...)
11:01:54 <chris> but I have another approach in mind that I'll do tomorrow, no flag required
11:02:08 <gjanssens> Sure, you're the guile expert, so be creative!
11:02:42 <gjanssens> It would be nice if it's transparent to calling code like net-worth and friends
11:02:48 <chris> of course
11:03:12 <gjanssens> :)
11:04:07 <chris> ok gtg now
11:05:06 * chris chris-away
11:05:09 *** chris is now known as chris-away
11:14:32 *** tienne has joined #gnucash
11:40:53 *** bertbob has quit IRC
11:41:56 *** bertbob has joined #gnucash
11:45:51 *** bertbob has quit IRC
11:47:12 *** bertbob has joined #gnucash
12:33:17 *** gjanssens has quit IRC
12:44:52 *** bertbob has quit IRC
12:46:10 *** nimish has quit IRC
12:53:45 *** bertbob has joined #gnucash
12:54:55 *** Agfarmer18 has joined #gnucash
13:04:01 *** joatmon has joined #gnucash
13:04:35 <joatmon> Hi, I have a question about autofill.
13:15:53 <joatmon> I'm on a Mac and when I start typing and gnucash autofills the description, my cursor stays in the position where I was typing but then if I keep typing the rest of the autofilled description stays after my cursor.
13:16:45 <joatmon> It's annoying to have to keep deleting everything that gets autofilled if the transaction I'm entering is different from the autofill.
13:17:45 <joatmon> I'm using version 3.2.
13:20:10 <joatmon> I just realized there's a later version out, I'll update and see if that fixes it.
13:21:17 *** kapil___ has quit IRC
13:37:06 *** fabior has quit IRC
13:39:24 *** Assador has joined #gnucash
13:39:28 <Assador> Http://for2too.fun please pm my. Female 25, 120-60-100. Blond. Search man.
13:44:13 <Assador> I love anal sex
13:44:18 *** nimish has joined #gnucash
13:47:20 *** Assador has quit IRC
14:00:00 *** nimish has quit IRC
14:18:50 <joatmon> I'm unable to open version 3.4 - I get a "CoreServicesUIAgent quit unexpectedly" error and have to force quit guncash.
14:38:41 <finster> joatmon: update to gnucash 3.4. It's solved there
14:39:20 <finster> Sorry, read too late you're getting an error with Gnuchash 3.4. What macOS version are you on?
14:39:27 <joatmon> I just tried but I can't open it - I get a "CoreServicesUIAgent quit unexpectedly" error
14:39:34 <joatmon> 10.9 Mavericks
14:39:47 <joatmon> I tried 3.3 and it opens fine
14:40:21 <finster> The devs might want to clarify any new dependecies introduced with version 3.4
14:40:53 <joatmon> Does it not work on 10.9?
14:42:09 *** Agfarmer18 has quit IRC
14:43:48 <joatmon> I just got the font down to a reasonable size in 3.3 and it seems to be fixed here too.
14:58:58 <joatmon> If anyone has any tips for getting 3.4 to verify so it will open, I'd be happy to hear them.
15:16:14 <joatmon> In case anybody cares, I got 3.4 to open. I did two things, not sure if either/both are related...
15:16:59 <joatmon> First I set Allow apps from anywhere in SysPrefs > Security & Privacy, General tab
15:17:18 <joatmon> I tried opening it and it just crashed silently.
15:17:54 <joatmon> Then I changed the folder name from Gnucash-3.4 to just Gnucash, and it opened fine.
15:18:59 <joatmon> I set both back to original settings and it still opens fine. Not sure what fixed it, but it works now. ¯\_(ツ)_/¯
15:19:20 <finster> joatmon: so it was a general security issue on osx then
15:19:25 <joatmon> Thanks for your help finster!
15:19:33 <joatmon> I'm guessing that's probably what it was.
15:20:00 <finster> joatmon: n.b.: you can bypass setting "Allow apps from anywhere" by just doing a right-click on the file, then open
15:20:05 <joatmon> Like I said, the first time after changing the security settings it crashed silently right away.
15:20:19 <joatmon> Yes, I had tried that too with no difference.
15:21:07 <joatmon> Not sure why changing the actual setting made it work.
15:21:37 <joatmon> I didn't make the changes scientifically though, so maybe I missed something. X-D
15:50:34 <jralls> joatmon: Please open terminal and run `spctl -av path/to/Gnucash-3.4.app`.
15:51:46 <joatmon> "a sealed resource is missing or invalid"
15:52:10 <jralls> Crud.
15:55:58 <joatmon> Interestingly, I get the same result on my 3.3 .app, which didn't have any problems.
15:56:35 *** bertbob has quit IRC
15:57:28 <joatmon> The 3.2 app came back accepted though.
15:59:14 *** boldstripe has joined #gnucash
16:01:57 *** finster has left #gnucash
16:05:46 *** sjenkins has quit IRC
16:29:38 *** bertbob has joined #gnucash
17:01:04 <chris-away> @tell gjanssens YearDelta also needs special handling... because report could choose start date 29/2/2016, inteval 1 year...
17:01:04 <gncbot> chris-away: The operation succeeded.
17:28:35 *** boldstripe has quit IRC
17:31:28 *** boldstripe has joined #gnucash
17:54:56 *** Agfarmer18 has joined #gnucash
18:02:36 *** boldstripe has quit IRC
18:07:45 *** jrp has joined #gnucash
18:07:45 *** jrp_ has quit IRC
18:14:39 *** Agfarmer18 has quit IRC
18:38:37 *** joatmon has quit IRC
19:07:15 *** jrp_ has joined #gnucash
19:08:58 *** jrp has quit IRC
19:14:35 *** trying2Simplify has joined #gnucash
19:15:36 *** gour has quit IRC
19:25:01 *** tienne has quit IRC
19:38:43 *** jervin has joined #gnucash
19:43:47 *** jervin has quit IRC
20:05:09 *** oozer has quit IRC
20:15:35 *** gncbot sets mode: +o fell
20:29:18 *** chris-away is now known as chris
20:29:39 <chris> @tell gjanssens pushed commit in PR, this is the best I can do for now
20:29:39 <gncbot> chris: The operation succeeded.
20:37:03 <warlord> .
20:42:07 *** trying2Simplify has quit IRC
20:49:17 *** jervin has joined #gnucash
20:57:38 *** jervin has quit IRC
21:02:29 <chris> @tel gjanssens https://imgur.com/a/GcFjnF9 on my branch #450
21:02:29 <gncbot> chris: Error: "tel" is not a valid command.
21:02:38 <chris> @tell gjanssens https://imgur.com/a/GcFjnF9 on my branch #450
21:02:38 <gncbot> chris: The operation succeeded.
21:10:12 *** bertbob has quit IRC
21:12:00 *** bertbob has joined #gnucash
21:15:55 *** bertbob has quit IRC
21:17:21 *** bertbob has joined #gnucash
21:24:07 *** badger92 has quit IRC
21:26:26 *** bertbob has quit IRC
21:27:57 *** bertbob has joined #gnucash
21:31:00 *** bertbob has quit IRC
21:40:31 *** jervin has joined #gnucash
21:44:00 <fell> chris: Nice, but how about the other direction. Unfortunatly my records start in february, but i don't want the 28. of each month but the last day of the month.
22:01:59 *** jervin has joined #gnucash
22:05:04 *** bertbob has joined #gnucash
22:05:31 <chris> sorry fell we had to make an executive decision... my advice is to start report 31-jan
22:06:12 *** chris has quit IRC
22:06:33 *** chris has joined #gnucash
22:09:52 <fell> chris: NP, but put it in the docs.
22:10:57 *** benoitg has quit IRC
22:18:34 *** bertbob has quit IRC
22:19:33 *** nimish has joined #gnucash
22:25:37 *** bertbob has joined #gnucash
22:26:50 *** jervin has quit IRC
22:29:32 *** bertbob has quit IRC
22:30:54 *** bertbob has joined #gnucash
23:17:19 *** chris has quit IRC
23:17:22 *** chris has joined #gnucash
23:19:09 *** joker has joined #gnucash
23:22:12 *** joker has quit IRC
23:22:57 *** joker has joined #gnucash
23:33:53 *** nimish_ has joined #gnucash
23:34:08 *** nimish has quit IRC
23:34:41 *** nimish_ is now known as nimish
23:41:32 *** joker has quit IRC