2020-07-14 GnuCash IRC logs
09:09:24 *** gncbot has joined #gnucash
09:20:02 *** jw4 has quit IRC
09:20:25 *** jw4 has joined #gnucash
09:20:25 *** ChanServ sets mode: +v jw4
09:20:30 *** hussam has joined #gnucash
09:20:30 *** ChanServ sets mode: +v hussam
09:28:53 *** ChanServ sets mode: +qo warlord warlord
09:28:55 *** warlord sets mode: +o gncbot
09:29:19 <warlord> So, my v4 network came back around 9:09am ET this morning. YAY. Now working to get v6 operational again.
09:33:02 *** lmat has quit IRC
09:33:06 *** lmat has joined #gnucash
09:37:50 <chris> ooh: https://i.imgur.com/ECdrz58.png http://sankeymatic.com/build/
09:41:24 *** Mechtilde has quit IRC
09:41:42 *** Mechtilde has joined #gnucash
09:44:04 <chris> Income->Asset/Liab->Expenses
09:44:06 <chris> of course
09:52:09 *** storyjesse has quit IRC
09:52:48 <warlord> And IPv6 is back. YAY
10:05:30 <chris> \o/
10:14:40 <warlord> Glad I remembered the bug in the original configuration ;)
10:22:12 *** sbluhm has quit IRC
10:33:24 *** Mechtilde has quit IRC
10:34:05 <chris> IMHO sankeymatic would be a very nice addition to charts
10:35:39 <chris> especially with animation, i.e. monthly income/expense over time
10:39:27 *** sbluhm has joined #gnucash
10:39:27 *** ChanServ sets mode: +v sbluhm
10:51:25 <chris> jralls_afk: I think we're ready to enable guile-3 support. shall I push, and await testers from other distro users?
11:02:22 *** jervin has joined #gnucash
11:05:15 <fell> chris, is "G_" added to the xgettext call in cmake?
11:06:34 <chris> check the PR
11:07:12 <fell> I am almost out of house.
11:07:43 *** jervin has quit IRC
11:16:05 *** kael has joined #gnucash
11:16:05 *** ChanServ sets mode: +v kael
11:26:32 *** ecdhe has quit IRC
11:28:14 *** sbluhm has quit IRC
11:32:18 <fell> Chris: another place to update: https://wiki.gnucash.org/wiki/Custom_Reports#Get_to_know_Gettext
11:33:03 *** ecdhe has joined #gnucash
11:37:11 *** fell has quit IRC
11:37:45 *** fell has joined #gnucash
11:37:45 *** ChanServ sets mode: +o fell
11:55:32 *** AdrienM has quit IRC
11:55:47 *** AdrienM has joined #gnucash
11:56:50 *** sbluhm has joined #gnucash
11:56:50 *** ChanServ sets mode: +v sbluhm
12:01:44 *** AdrienM has quit IRC
12:01:55 *** AdrienM has joined #gnucash
12:02:36 *** ChanServ sets mode: +v AdrienM
12:04:18 *** sbluhm has quit IRC
12:07:07 *** lmat has quit IRC
12:08:36 *** ramontjunior has quit IRC
12:12:29 *** nullquery has joined #gnucash
12:12:53 *** ChanServ sets mode: +v nullquery
12:13:16 <nullquery> Hi.
12:13:27 <nullquery> Er, I shouldn't open like that.
12:14:59 <nullquery> I feel like a fool, which is why I'm asking here first instead of posting this elsewhere, but... I'm running my own small business in the UK, but I have EU clients. They pay me in EUR and I receive in GBP. My base currency is GBP. I've had to use GNUCash's Trading Accounts feature to deal with conversion of EUR->GBP.
12:16:02 <warlord> nullquery, do you need to actually know how much they paid you in EUR?
12:16:09 <nullquery> My problem is with the balance sheet. It shows "Trading Losses" which seems to be a calculated value. However, this value is 1 penny short from the Total Equity. I would very much like to correct this, but I'm not sure how this value is calculated.
12:16:53 <nullquery> @warlord Yes, I think so. I'm invoicing them in EUR, so while the invoice is outstanding I will have Accounts Receivable asset in EUR, which is converted to GBP upon payment.
12:16:53 <gncbot> nullquery: Error: "warlord" is not a valid command.
12:17:35 * nullquery starts to nervously sweat. "W-what do you mean, gncbot?"
12:18:18 <warlord> nullquery, you put an '@' in front of "warlord". That @ tells gncbot that what follows is a command to gncbot, and "warlord" is not a valid comment.
12:18:22 <warlord> command, even
12:18:36 <nullquery> Oh. Sorry, gncbot! :)
12:18:49 *** lmat has joined #gnucash
12:18:57 <warlord> Anyways, when you have a change in exchange rate, you could have a loss. Most likely the 1 penny is a rounding error.
12:19:13 *** jralls_afk is now known as jralls
12:19:53 <nullquery> That's what I was thinking. I would like to correct the rounding error and have the correct value show up in the Balance Sheet, but I'm not sure how to do that without influencing the value of the Total Equity.
12:20:37 <chris> nullquery: which version are you using?
12:20:44 <nullquery> Or alternatively, implementing a patch in the code to correct this on the balance sheet. The total of my equity account - Total Equity on the balance sheet = Trading Losses
12:20:57 <nullquery> Build ID: Version 4.0+ (2020-06-27)
12:21:23 <chris> and which Balance Sheet are you using
12:21:28 <nullquery> I updated to what I think is the latest version and tried with that, but there doesn't seem to be an option for this and I was just driving myself crazy with it
12:21:42 <nullquery> Reports -> Assets & Liabilities -> Balance Sheet
12:23:23 <chris> Immediately I won't have an immediate answer for you, except this Trading Accounts issue runs very deep within mathematics of multi-currency accounting :)
12:24:16 <chris> 0.01 is a rounding error there are some recent changes to reduce rounding in the internal calculations
12:24:30 <chris> ^ e.g. gnc:make-commodity-collector
12:26:12 <chris> I think a CPA/ACCA will advise you to record all your business income converted into your local currency irrespective of the invoice contents
12:27:23 <chris> (I deal with mostly the reports) but jralls is probably more capable of offering a more user-friendly suggestion :)
12:30:02 <nullquery> When I send an invoice I record it as an asset in Accounts Receivable in EUR. This is converted by the balance sheet to GBP at the exchange rate of my accounting end period.
12:30:46 <nullquery> When I get paid, I don't know the exchange rate that was used to receive the money (bank informs me for some clients, but not for others) and so I have GNUCash calculate the exchange rate. Definitely possible to have rounding errors there...
12:32:18 <nullquery> I wonder where the code is that calculates the Trading Gains/Losses
12:32:22 <jralls> nullquery, you won't get rounding errors if you always enter the amount in GBP and let let GnuCash compute the rate. Rounding errors happen when you enter the rate and let GnuCash compute the amount.
12:33:37 <nullquery> That happened originally when I worked with the exchange rates. I thought it might be easier to just enter the amount that I actually got in the bank and let GNUCash deal with it.
12:34:14 <nullquery> I'm concerned though... What if I have to enter 2 EUR->GBP transactions on the same day? I saw that the price database keeps records per date, would it handle this correctly?
12:34:42 <jralls> Not only easier, better because amounts are rounded to the currency's smallest unit (e.g. pence) and rates are exact rational numbers.
12:35:27 <nullquery> Yay! I did good! :D
12:36:53 <jralls> The price database is separate from the per-transaction calculation. And BTW you want to use "average cost" price source when running your balance sheet so that you get the effective exchange rate for the transactions instead of one from the pricedb.
12:37:44 <nullquery> Shouldn't I use "nearest in time" so that the exchange rates are calculated vs. the exchange rate for the end of my accounting period?
12:38:20 <jralls> This works quite well as long as all of the transactions are composed of splits in your home currency and one other. It fails for transactions that are completely in other currencies, see https://bugs.gnucash.org/show_bug.cgi?id=797796.
12:40:13 <jralls> "Nearest in time" uses the rate from the pricedb closest to the date of the report. That will inevitably produce imbalances that will show up as unrealized or unbooked gains/losses.
12:43:29 <jralls> That said, you do have trading gains and losses: When you invoice someone in EUR for something that you had in inventory as GBP there's an exchange; when they pay you in EUR and the bank converts it back to GBP there's another exchange. If the rates of the two exchanges differ then there's a trading G/L that you need to book. Your trading accounts should make that apparent.
12:45:50 <jralls> As long as you always enter amounts and let GnuCash handle the rates then there should be no rounding errors and if you've booked the trading gains correctly both the balance sheet and trial balance report should balance when using "average cost".
12:46:51 <nullquery> In my case, I'm exchanging a service so I don't have to worry about the first exchange. However, I do worry about the second- if I do "average cost" I get a trading loss/gain of 0.00. The "nearest in time" calculation makes the most sense, I think, because I'd be reporting what I have in GBP *now* instead of back then?
12:47:52 *** jralls_afk has joined #gnucash
12:47:53 *** ChanServ sets mode: +o jralls_afk
12:48:12 <chris> aside: jralls: are we ready to enable guile-3? It WFM; shouldn't affect ppl on guile-2.2 only...
12:48:49 *** jralls has quit IRC
12:49:06 <chris> nullquery: nearest-in-time means foreign currencies are converted to report-currency using foreign->home pricedb.
12:49:15 <jralls_afk> Bleeping Comcast. Depends. If you want to know what your EUR AP is worth in GBP at today's rate then put that rate in the pricedb and run the balance sheet with nearest in time.
12:49:17 <chris> this would make sense it you're holding an EUR balance?
12:50:19 <jralls_afk> chris, I need to test it on my OpenSuSE VM with Guile3.
12:50:42 *** guak has joined #gnucash
12:50:46 *** jralls_afk is now known as jralls
12:50:47 <chris> sure, and let me know any errors. I'll push a copy-and-paste bug then.
13:22:53 *** sbluhm has joined #gnucash
13:22:53 *** ChanServ sets mode: +v sbluhm
13:23:10 *** Mechtilde has joined #gnucash
13:30:31 *** CDB-Work has joined #gnucash
13:30:31 *** ChanServ sets mode: +v CDB-Work
13:32:22 *** David has quit IRC
13:32:36 *** David has joined #gnucash
13:48:23 *** sbluhm has quit IRC
13:49:17 <jralls> chris, no problems, merged & pushed.
14:03:44 <fell> chris, guile 3 is not mentioned in https://wiki.gnucash.org/wiki/index.php?title=Custom_Reports&diff=next&oldid=16568#Where_to_find_existing_reports
14:06:23 <fell> Perhaps you should search the wiki at least for "guile", optionally "Scheme", "scm", "report" and review that pages.
14:07:22 <fell> later apply the same on the docs. ;-)
14:09:08 *** sbluhm has joined #gnucash
14:09:08 *** ChanServ sets mode: +v sbluhm
14:10:55 *** frakturfreak has joined #gnucash
14:10:55 *** ChanServ sets mode: +v frakturfreak
14:11:00 *** Tony has joined #gnucash
14:11:20 <jralls> fell, It looks to me that where to find existing reports is still correct: The guile version is still typically 2.2 or 2.0. OpenSuSE is atypical.
14:12:21 <fell> Do we not change the directory <guile-version*>
14:12:32 <fell> for 3
14:13:29 <jralls> No. That can be 2.0, 2.2, or 3.0. There's a note underneath the table explaining that.
14:24:10 <fell> There are no Flatpaks after 2020-07-12. Related to the network outage?
14:24:40 <warlord> fell, yes
14:25:08 <warlord> I expect new ones will build tonight, now that code actually has network access again.
14:26:10 <fell> Do you really believe the connection will last for longer? ;-) Can we restart them?
14:26:11 *** suukim has quit IRC
14:27:34 <warlord> Are you asking if I think the network is going to go down again? No, I do not think it is going to go down again -- at least not immediately. I'm sure it will, eventually.
14:27:46 <warlord> As for kicking the build manually, does 12 hours really matter?
14:28:41 <fell> I am already waiting 2 days for the new aqbanking stuff.
14:35:01 <warlord> fell, okay, I fired it off by hand.
14:36:12 <fell> Thanks!
14:41:30 <fell> I am wondering, where the difference in the setup lies. The Windows builds succeeded, but Flatpak failed.
14:49:56 <jralls> Flatpak didn't fail, it didn't even start.
14:56:54 <warlord> The difference is that the windows VM is on a private network, whereas code is on my public network. It was only my public network that went out, so code failed (as did all my other servers on the public network), but stuff on the private network continued to work... And the private network could reach the public network (it's all internal, after all).
14:57:00 <warlord> So the windows builds worked.
15:09:16 <fell> jralls, it failed: https://code.gnucash.org/builds/flatpak/build-logs//?C=M;O=D
15:09:55 <fell> early failing logs are not moved in the right directory.
15:12:47 *** jervin has joined #gnucash
15:12:53 <jralls> Oh, that's interesting. The build failed too: Error: module gwenhywfar: Child process exited with code 1
15:17:16 <warlord> Let me know if you want me to kick off another.
15:19:41 <jralls> Doing a build on my F31 VM first to get more info about what broke.
15:25:29 <warlord> ok
15:30:47 *** jervin has quit IRC
15:32:16 *** jervin has joined #gnucash
15:46:37 *** David1 has joined #gnucash
15:47:31 *** David has quit IRC
15:47:31 *** David1 is now known as David
15:51:49 *** jervin has quit IRC
15:53:52 *** jervin has joined #gnucash
16:03:31 *** jervin has quit IRC
16:04:09 *** Mechtilde has quit IRC
16:06:39 *** David has quit IRC
16:06:43 *** David has joined #gnucash
16:09:36 <jralls> warlord, it built fine for me. The line above the child process error says "Unable to allocate instance id" and that seems to be a flatpak issue. At least when I google it all of the answers seem to be related to flatpak. Unfortunately none of them offer any insight.
16:16:35 *** sbluhm has quit IRC
16:21:48 <fell> I had committed afer I had built successful under opensuse tumbleweed.
16:21:48 <fell> which flatpak version is in te buildserverß
16:21:48 <fell> c/ß/?/
16:23:07 <heredoc> Russian sed? :D
16:29:38 <jralls> перл
16:30:05 <warlord> fell, flatpak-builder-1.0.7-1.fc29.x86_64
16:39:07 <fell> I have 1.0.9-1.3
16:40:28 <fell> Perhaps we need to upgrade fedora to 30 or 31?
16:40:41 <warlord> Code is running F29. Retest there.
16:40:55 <warlord> Upgrading code is not something I have time to do in the short-term.
16:43:26 <fell> 29 is unsupported since last year
16:44:34 <fell> Recent is 32
16:45:32 <warlord> fell, your point?
16:45:39 <warlord> that's what code is running
16:46:36 <fell> I just looked for the current state.
16:47:04 <warlord> I tend to upgrade on a 1.5.2.5-year cycle.
16:47:18 <fell> But if there is a lfatpak-builder bug and you get no lnger updates ...
16:47:42 <warlord> Then don't use newer features. It was building fine last week.
16:49:23 <warlord> Indeed, it was building fine 2 days ago.
16:50:59 *** lmat has quit IRC
16:52:24 <fell> In between I only changed the versions of aqbanking and gwenhywfar
16:52:27 <warlord> OH, I think I remember what the instance id issue is..
16:54:01 *** lmat has joined #gnucash
16:54:27 <warlord> Yep, that was the issue. Flatpak it building.
16:54:55 <warlord> For posterity, if you get
16:54:56 <warlord> error: Unable to allocate instance id
16:55:12 <warlord> the issue is that when you su(do), you need to access a login shell
16:55:27 <warlord> so you can't su -s /bin/bash gnucash-docs
16:55:34 <warlord> you need to su -l -s /bin/bash gnucash-docs
16:56:04 <warlord> It works fine from CRON because it sets the full env, but when you su (without -l) then the environment is "weird" and flatpak gets confused.
16:57:13 *** lmat has quit IRC
17:04:24 *** lmat has joined #gnucash
17:07:22 *** User has quit IRC
17:09:50 *** frakturfreak has quit IRC
17:42:09 *** jervin has joined #gnucash
17:45:30 *** ecdhe has quit IRC
17:57:42 <warlord> fell, looks like the maint build finished; I suspect master will shortly (if not already done)
18:01:30 <fell> Thanks, warlord! Then I can write an announcement.
18:07:33 <warlord> Okay
18:13:23 *** lmat has quit IRC
18:15:43 *** lmat has joined #gnucash
18:42:16 *** dnbNcoffee has quit IRC
18:46:15 *** GlacialSpain has joined #gnucash
18:49:13 *** dnbNcoffee has joined #gnucash
18:54:41 *** Aussie_matt has joined #gnucash
19:01:57 *** GlacialSpain has quit IRC
19:12:39 *** iReflex has joined #gnucash
19:12:39 *** ChanServ sets mode: +v iReflex
19:24:42 *** Agfarmer18 has joined #gnucash
19:32:19 *** kael has quit IRC
19:49:10 *** shaggy has quit IRC
19:49:38 *** shaggy has joined #gnucash
20:07:14 *** guak has quit IRC
20:12:45 <chris> jralls: yay for guile-3! It may be nice to make windows build carry guile-3 instead of guile-2.2 too :)
20:33:37 *** jlesh has joined #gnucash
20:35:53 *** jlesh is now known as jmlesher
20:38:24 *** jmlesher is now known as jlesher
20:40:22 <CDB-Work> chris: FYI, i wanted to work more on the bug sunday, but then bugzilla wad down :( -- now i wont have time until the weekend
20:40:35 *** ChanServ sets mode: +v jlesher
20:56:25 *** jlesher has left #gnucash
21:27:46 <chris> ok CDB-Work. no rush.
21:27:49 *** chris has left #gnucash
21:32:07 *** jervin has quit IRC
21:37:09 *** jervin has joined #gnucash
21:49:23 *** jervin has quit IRC
22:26:26 *** hussam has quit IRC
22:35:42 *** CDB-Work has quit IRC
22:42:45 *** dnbNcoffee has quit IRC
22:51:28 *** fell has quit IRC
22:51:47 *** fell has joined #gnucash
22:51:48 *** ChanServ sets mode: +o fell
23:52:28 *** chris has joined #gnucash
23:52:28 *** ChanServ sets mode: +v chris