2022-05-13 GnuCash IRC logs

01:11:28 *** warlord has quit IRC
01:11:33 *** warlord has joined #gnucash
01:28:11 *** sbluhm has joined #gnucash
01:32:37 *** warlord2 has joined #gnucash
01:32:37 *** warlord has quit IRC
01:32:40 *** gncbot has joined #gnucash
01:36:16 *** gjanssens has joined #gnucash
01:36:16 *** ChanServ sets mode: +o gjanssens
01:45:58 *** warlord_ has joined #gnucash
01:46:44 *** warlord2 has quit IRC
01:49:11 *** sbluhm has quit IRC
01:49:46 *** fell has quit IRC
01:51:06 *** fell has joined #gnucash
01:51:06 *** ChanServ sets mode: +o fell
02:25:16 <chris> hmm there's duplicate "New Invoice..." and "New Invoice" buttons in invoice-editor
03:21:01 *** sbluhm has joined #gnucash
03:35:27 <chris> gjanssens: is it a fair assumption that all TXN_TYPE_LINK will have all splits belonging to the APAR account, and all splits will be attached to a document lot?
03:35:54 <chris> (ref #1201)
03:36:05 <chris> ^ or warlord
03:36:41 <chris> if yes, then the LINK txns will be converted to TXN_TYPE_PAYMENT by #1201 and must be hidden from new-owner-report
03:48:46 *** sbluhm has quit IRC
04:11:53 *** sbluhm has joined #gnucash
04:19:23 *** sbluhm has quit IRC
04:34:04 *** AdrienM has quit IRC
04:34:49 *** AdrienM has joined #gnucash
04:34:50 *** ChanServ sets mode: +v AdrienM
05:02:42 *** storyjesse has joined #gnucash
06:03:20 *** storyjesse has quit IRC
06:17:30 *** sbluhm has joined #gnucash
06:40:22 *** jmdaemon has joined #gnucash
06:43:46 *** sbluhm has quit IRC
07:00:42 *** sbluhm has joined #gnucash
07:10:43 *** Aussie_matt has joined #gnucash
07:15:23 *** jmdaemon has quit IRC
07:34:28 *** sbluhm has quit IRC
07:42:06 *** warlord_ is now known as warlord
07:42:07 *** ChanServ sets mode: +qo warlord warlord
07:42:56 <warlord> chris, I am not aware of what LINK txns are. Those were added later, I think.
07:43:05 *** warlord sets mode: +o gncbot
07:58:21 *** chris_ has joined #gnucash
07:58:22 *** gncbot sets mode: +o chris_
07:59:55 *** chris has quit IRC
08:20:53 *** paul has joined #gnucash
08:52:03 *** sbluhm has joined #gnucash
08:57:08 *** chris_ is now known as chris
08:57:36 <chris> warlord: ok I think they're gjanssens work; aims to marry up Invoices to CNs
08:58:09 <warlord> Right.
08:59:11 *** jonp has quit IRC
09:00:52 *** jonp has joined #gnucash
09:19:03 *** sbluhm has quit IRC
09:33:25 <gjanssens> chris: yes they are uniquely used when two document lots need to be linked together
09:33:48 *** sbluhm has joined #gnucash
09:34:02 <gjanssens> And I agree this could have been done with TXN_TYPE_PAYMENT as well. But I only figured that out in hindsight.
09:34:36 <gjanssens> In fact a TXN_TYPE_LINK is a special case of TXN_TYPE_PAYMENT, nae
09:34:40 <gjanssens> oops
09:35:21 <gjanssens> namely another document is used as "payment method" instead of cash or bank.
09:35:24 <warlord> hindsight is always 20/20
09:35:46 <gjanssens> But technically you can just use TXN_TYPE_PAYMENT everywhere.
09:36:41 <gjanssens> I don't know offhand if there's any logic that specifically depends on TXN_TYPE_LINK exclusively. If so that may have to be revisited of course.
09:36:47 <gjanssens> warlord: indeed :)
09:37:03 *** jonp has quit IRC
09:38:38 <chris> well the gist of #1201 is- xaccTransGetTxntype will read the splits and determine (and cache) the TxnType instead of reading from KVP. it doesn't return TXN_TYPE_LINK. if my assumption is correct
09:38:56 <chris> (i.e. LINK txns are always ReadOnly, and have no non-APAR splits) then #1201 works well.
09:43:45 <chris> if #1201 then there's no more Txn_Type_LINK anymore
09:46:37 <chris> thus datafiles damaged by clicking on type column in APAR register will be silently 'fixed'
09:50:21 <chris> P accidentally changed to I will be silently reverted to P in the register. but the kvp txn-type won't be automatically removed.
10:02:19 *** sbluhm has quit IRC
10:07:10 *** sbluhm has joined #gnucash
10:12:10 *** lmat has quit IRC
10:19:10 *** sbluhm has quit IRC
10:21:24 *** lmat has joined #gnucash
11:09:12 *** Aussie_matt has quit IRC
11:18:48 *** bertbob has quit IRC
11:21:05 *** bertbob has joined #gnucash
11:21:05 *** ChanServ sets mode: +v bertbob
11:56:47 *** Vafa has joined #gnucash
11:56:47 *** ChanServ sets mode: +v Vafa
12:12:58 *** guak has joined #gnucash
13:04:21 <fell> I am not sure that "silently fixing" is really a good idea.
13:05:04 <fell> Imagine I open a file in a new vesion only to search something.
13:06:15 <fell> The on closing, I am wondering, why it wants to save the file.
13:06:54 <fell> I think, I might accidently changed something and reject saving.
13:07:23 <fell> On the next start the repair has to be done again.
13:08:47 <fell> IIRC that happens on all my older test files—again and again,
13:10:45 <fell> Perhaps we should set a "Saving reason": User changed something or program fixed something.
13:31:01 *** sbluhm has joined #gnucash
13:39:12 *** AdrienM has quit IRC
13:39:37 *** AdrienM has joined #gnucash
13:39:37 *** ChanServ sets mode: +v AdrienM
13:51:01 <jralls_afk> fell, added to the message in the save-on-close dialog?
13:51:43 <fell> Yes
13:51:44 <jralls_afk> OTOH IIUC chris's change won't write to the file, it will just ignore what's there and do what it thinks is right.
13:53:04 *** jralls_afk is now known as jralls
14:16:01 *** sbluhm has quit IRC
14:31:46 *** sbluhm has joined #gnucash
14:40:58 *** jmdaemon has joined #gnucash
15:56:31 *** sbluhm has quit IRC
16:31:42 *** gjanssens has quit IRC
18:40:31 *** Vafa has quit IRC
19:05:19 *** field^Mop has joined #gnucash
19:23:19 *** field^Mop has quit IRC
20:08:30 *** guak has quit IRC
22:02:00 <chris> IMHO #1201 the xaccTransSetTxnType should still be allow to write to kvp; it means old builds can read newer datafiles