2024-04-26 GnuCash IRC logs
00:16:20 *** jmck has joined #gnucash
00:19:20 *** jmck has quit IRC
01:09:22 <Antisoche> jrails: I did not sell incomplete fractions. I imported data from a CSV and the number that came in from there was rounded.
01:11:20 <Antisoche> The CSV importer doesn't work well reading transactions in splits but that's a whole other story... Hopefully the newer releases are better. I gather I'll be prompted to upgrade sometime around June.
01:12:21 *** jmck has joined #gnucash
01:15:21 *** jmck has quit IRC
01:41:34 *** fell has quit IRC
01:42:52 *** fell has joined #gnucash
01:42:53 *** ChanServ sets mode: +o fell
02:08:01 *** jmck has joined #gnucash
02:11:02 *** jmck has quit IRC
02:39:16 *** jmck has joined #gnucash
02:42:16 *** jmck has quit IRC
03:04:30 *** gjanssens has joined #gnucash
03:04:30 *** ChanServ sets mode: +o gjanssens
03:24:34 <Simon> simply stating that you expect newer versions to be better isn't going to change anything
03:25:03 <Simon> if you can provide a sample of the CSV lines you're importing it would be possible to comment further
03:25:41 <Simon> in my experience things like CSV exports vary from wrong to totally wrong in areas like unit counts
03:27:30 <Simon> if you're using Ubuntu, try simplescreenrecorder to record what you're doing
03:40:02 *** jmck has joined #gnucash
03:43:02 *** jmck has quit IRC
04:36:50 *** jmck has joined #gnucash
04:39:50 *** jmck has quit IRC
04:40:48 *** Kaligula has joined #gnucash
04:41:20 *** jmck has joined #gnucash
04:42:23 *** ChanServ sets mode: +v Kaligula
04:44:21 *** jmck has quit IRC
05:12:37 *** jmck has joined #gnucash
05:15:37 *** jmck has quit IRC
05:41:50 *** jmck has joined #gnucash
05:44:50 *** jmck has quit IRC
05:58:09 *** jmck has joined #gnucash
06:01:10 *** jmck has quit IRC
06:43:07 *** jmck has joined #gnucash
06:46:07 *** jmck has quit IRC
06:52:41 *** jmck has joined #gnucash
07:24:26 *** chris has joined #gnucash
07:24:26 *** ChanServ sets mode: +v chris
07:24:26 *** gncbot sets mode: +o chris
07:40:09 <warlord> .
08:42:56 *** jmck has quit IRC
08:53:51 *** chris has quit IRC
09:19:11 *** chris has joined #gnucash
09:19:12 *** ChanServ sets mode: +v chris
09:19:12 *** gncbot sets mode: +o chris
09:28:03 *** jmck has joined #gnucash
09:31:03 *** jmck has quit IRC
10:07:58 *** warlord has quit IRC
11:17:51 *** jmck has joined #gnucash
11:20:51 *** jmck has quit IRC
11:21:14 *** jmck has joined #gnucash
11:30:06 *** warlord has joined #gnucash
11:30:06 *** gncbot sets mode: +o warlord
11:31:18 *** jmck has quit IRC
11:56:44 *** chris has quit IRC
12:45:32 *** jmck has joined #gnucash
12:48:32 *** jmck has quit IRC
13:46:38 *** jmck has joined #gnucash
13:49:38 *** jmck has quit IRC
14:40:27 <Antisoche> Simon: I'm running v4.8 and I recall speficially mentions of changes to the CSV importer when I was looking at changelogs. At the moment I just need to get this done and I'm making progress so it's okay for me to wait until the new version comes in the Ubuntu LTS update. Just FYI: my issue is importing data from Fidelity. They have the to/from on different lines and the "multi-split" option in the importer doesn't recognize what it's doing, so 1/2 the
14:40:29 <Antisoche> imported txs are "0" values. Then I have to manually assign the split to an income:dividend account, which I may have to do anyway. It loses the "Price" value. I haven't thought about if that makes sense, but it does make importing data tedious because I end up doing ~60% of the data input by hand.
14:48:07 *** jmck has joined #gnucash
14:51:08 *** jmck has quit IRC
14:51:48 <Simon> you keep expecting that things will be different in a later version
14:52:34 <Simon> if there is a specific scenario that you think is a bug you should search for it and/or raise it on the bug tracker
14:53:29 <Simon> for the case of Fidelity, in my experience they only provide truncated units but your country/system may be different
14:58:01 *** CDB-Man has quit IRC
15:07:42 <Antisoche> Fair, but I don't want to file a bug if something has already changed. That just makes stuff harder for everyone. If I have the same problem in the new version I'll consider filing a bug. The units issue has been driving me crazy, but "Quantity Purchased" is always an estimate / is rounded - the current price and amount credited/spent are the only values I rely on. Fortunately I only have few transactions. Will see what ETrade does later...
15:12:49 *** CDB-Man has joined #gnucash
15:12:49 *** ChanServ sets mode: +v CDB-Man
15:25:32 <Simon> you can't work like that, it's not going to match when you try to reconcile it
15:26:28 <Simon> if they won't tell you the exact units you actually own I'd try to move to somewhere else that will
15:26:58 <Simon> I'm surprised they think they can do that for shares
15:28:09 <Simon> I have a script to obtain the unit count the day before and on the day of a change to work out what quantity of units to record
15:28:33 <Antisoche> In the CSV the "Quantity" field is limited to 3 fractional digits. 0.6666... is reported as 0.667. I don't think this is uncommon.
15:29:10 <Simon> it's not going to work if you actually have a different quantity and you try to reconcile your balance with theirs
15:30:17 <Simon> I have one provider that gives me a "Quantity Ordered" to 4 decimal places (this one is correct) and "Quantity Filled" to 2 decimal places (this one is rounded) which is crazy
15:30:43 <Simon> but at least one of the numbers is correct
15:30:58 <Simon> Fidelity are rounding the display of everything everywhere to 2 decimal places
15:31:08 <Simon> when in reality it's 6 decimal places internally
15:32:28 <Simon> you can approximate it with the price but at some point it's going to diverge, and can't match when they're rounding something up
15:32:39 <Antisoche> I'm not sure what the answer is but I suspect it's in the fraction share setting of the security. I have it set at 1/100000 and they have it as 1/1000, apparently. So my number is off by tenths of thousands. In my case I'm not buying, I'm getting a dividend that's reinvested and giving me small fractions of shares. The 2-decimal places problem I was having was GnuCash not letting me enter more than 0.02 fractions in the security when entering it on the
15:32:41 <Antisoche> CURRENCY trading account.
15:36:16 <Antisoche> When I zero-out an account I put in the necessary transaction amount and let the price float. I haven't looked closely but it's got to be fractions of pennies. The worst part so far is the "Fee" column, which is recorded as, eg, "0.22" but is really 0.216789 _per_share_.
15:36:26 *** Kaligula has quit IRC
15:37:37 <Simon> I can't find a recent bug for what you're describing
15:38:33 <Simon> yes, the value of the differences involved will be beyond trivial but it won't match
15:40:36 <Antisoche> grmble... 4.904 changelog has "Fix CSV price import file loading", "CsvImport - actually do as the warning indicates" and release for 5.0 says "Extensive changes to the CSV importer, resolving most known bugs"
15:41:49 <Antisoche> https://www.gnucash.org/news.phtml
15:48:53 *** jmck has joined #gnucash
15:51:53 *** jmck has quit IRC
17:13:06 *** gjanssens has quit IRC
17:23:32 *** jmck has joined #gnucash
17:26:33 *** jmck has quit IRC
17:45:16 *** jmck has joined #gnucash
19:12:00 *** chris has joined #gnucash
19:12:01 *** ChanServ sets mode: +v chris
19:12:01 *** gncbot sets mode: +o chris
19:40:43 *** jmck has quit IRC
19:41:05 *** jmck has joined #gnucash
19:42:40 *** jmck has quit IRC
19:42:42 *** jmck has joined #gnucash
20:00:02 *** raghukamath has quit IRC
20:11:56 *** chris has quit IRC
20:14:05 *** raghukamath has joined #gnucash
20:14:05 *** ChanServ sets mode: +v raghukamath
20:37:30 *** jmck has quit IRC
20:39:59 *** jmck has joined #gnucash
21:14:46 *** chris has joined #gnucash
21:14:46 *** ChanServ sets mode: +v chris
21:14:46 *** gncbot sets mode: +o chris
23:13:42 *** jmck has quit IRC
23:18:43 *** chris has quit IRC
23:29:11 *** chris has joined #gnucash
23:29:11 *** ChanServ sets mode: +v chris
23:29:11 *** gncbot sets mode: +o chris