2020-12-04 GnuCash IRC logs

00:32:07 *** hussam has joined #gnucash
00:32:07 *** ChanServ sets mode: +v hussam
00:38:55 *** fell has quit IRC
00:40:54 *** fell has joined #gnucash
00:40:54 *** ChanServ sets mode: +o fell
00:44:34 *** o01eg__ has quit IRC
00:45:40 <geneAum> CDB-Work: thanks so much for your thoughts around this. i'm not a major short trader, but i'm aiming to be, and i'm trying to make sure i grasp the accounting before i try to scale up, so your comments are great food for thought.
00:57:18 <CDB-Work> well I hope you realize that most shorts lose money
00:57:38 <CDB-Work> I would rather buy puts than open actual shorts
01:01:55 <geneAum> they're very short-lived positions, and the liability is only to myself (a "strategy" with a liability to a "treasury"), and i won't be TOO concerned if i end up giving up on the repurchase entirely, but i'd like to account for the positions properly
01:04:16 *** fell_laptop has joined #gnucash
01:04:16 *** ChanServ sets mode: +o fell_laptop
01:04:40 <CDB-Work> DR Cash (proceeds of the short), CR a STOCK account (negative asset) or a vanilla liability account
01:04:51 <CDB-Work> you won't be able to use mark to market price DB if you do not use STOCK accounts
01:05:07 <CDB-Work> I would rather do a topside adjustment for the short position, than lose out on mark to market pricing
01:05:42 <CDB-Work> just don't short TSLA and end up having to cover the short at a 30% loss as a result of one of Elon's market-moving tweets
01:06:00 <CDB-Work> I still think PUT options are a much safer play
01:06:37 *** fell has quit IRC
01:11:04 <geneAum> i've got a stockpile of an asset with a liquid spot market (but no derivatives market) and i'm successfully running a medium-frequency, sell-high/repurchase-low strategy, generating some pretty good returns from some otherwise idle assets, which i like holding but am not married to. i'm not REALLY shorting anything, just using shorts as a model.
01:18:45 *** frakturfreak has quit IRC
01:23:39 <CDB-Work> > i'm successfully running a medium-frequency, sell-high/repurchase-low strategy,
01:23:44 <CDB-Work> that's what we call "swing trading"
01:23:55 <CDB-Work> you are trading the normal cyclical swings in the market
01:24:17 <CDB-Work> most often, swing trading is done by a chartist, i.e. someone that trades purely on technical indicators such as MACD, RSI, etc
01:29:07 <geneAum> didn't know that's what i'm doing, good to know the label for it, thanks. this spot market is just more volatile than my transaction fee margins (i guess "it swings wide"), so i just trade as often as possible, not on any indicators
01:33:12 *** frakturfreak has joined #gnucash
01:38:11 *** sbluhm has joined #gnucash
01:38:11 *** ChanServ sets mode: +v sbluhm
01:39:17 *** fell_laptop is now known as fell
01:46:57 *** fell has quit IRC
01:49:58 *** fell has joined #gnucash
01:49:58 *** ChanServ sets mode: +o fell
02:07:15 *** Mechtilde has quit IRC
02:09:44 *** jervin has quit IRC
02:30:54 *** CDB-Work has quit IRC
03:20:04 *** pim has joined #gnucash
03:26:12 *** Mechtilde has joined #gnucash
03:49:05 *** gjanssens has joined #gnucash
03:49:05 *** ChanServ sets mode: +o gjanssens
04:19:58 *** chris has quit IRC
04:20:54 *** chris has joined #gnucash
04:20:54 *** ChanServ sets mode: +v chris
04:20:58 *** gncbot sets mode: +o chris
04:26:05 *** bertbob has quit IRC
04:30:10 *** bertbob has joined #gnucash
04:30:10 *** ChanServ sets mode: +v bertbob
04:50:01 *** User_ has joined #gnucash
04:51:04 *** hussam has quit IRC
05:00:42 *** Aussie_matt has quit IRC
05:28:21 <gjanssens> .
05:34:40 *** Aussie_matt has joined #gnucash
06:07:32 *** Aussie_matt has quit IRC
06:13:39 *** Herbert has joined #gnucash
06:13:46 *** Herbert has quit IRC
06:13:47 *** Agfarmer18 has joined #gnucash
07:20:59 <chris> fell: would you mind trying - instead of (define-syntax-rule (N_ x) x) modify to (define (N_x) x) in core-utils.scm for me and try make pot?
07:28:09 <chris> gjanssens: guix has an 'interesting' feature - if I insert a guix.scm into project root, it offers an alternate development environment to hack code. on guix, the simple command 'guix environment -l guix.scm' will load all dependencies to hack gnucash. are there equivalents eg in docker?
07:28:58 <chris> sample: https://pastebin.com/raw/vs15BWga
07:29:17 *** Agfarmer18 has quit IRC
07:36:00 <chris> fell: it's (define (N_ x) x) of course
07:37:18 <warlord> .
08:00:51 *** giuseppef has joined #gnucash
08:00:52 *** ChanServ sets mode: +v giuseppef
08:31:31 *** Mechtilde has quit IRC
08:35:11 <chris> fell: I've pushed a few cleanup commits. if 82d3bcd39 causes problems with 'make pot' let me know.
08:37:49 *** hussam has joined #gnucash
08:37:49 *** ChanServ sets mode: +v hussam
08:38:38 *** hussam has quit IRC
08:38:38 *** Mechtilde has joined #gnucash
08:39:29 *** hussam has joined #gnucash
08:39:29 *** ChanServ sets mode: +v hussam
08:41:31 *** giuseppef has quit IRC
08:48:25 *** Jimraehl1 has joined #gnucash
08:54:07 <gjanssens> chris: the syntax-rule change seems to work fine
08:54:24 <gjanssens> The same strings are extracted by make pot
08:55:13 <gjanssens> As for the dev environment, I don't know of such a feature relevant to gnucash
08:59:50 <chris> Ok. I trust you can understand the motivation... would automate the dependency installation, cmake and ninja builds. it's more opaque to the basic developer, but I can confirm it's OSS all the way
09:00:12 <chris> and technically can be triggered from a bare linux+guix installation
09:01:41 <gjanssens> chris: I didn't see your sample before I answered. I did now
09:02:13 <gjanssens> It's no different from a dockerfile that describes all the components that have to be installed and a runtime script.
09:02:34 <gjanssens> So if you are so inclined yes, you can create something similar for docker.
09:02:48 <gjanssens> I thought more magic was involved from your description :)
09:03:33 <chris> oh there's no magic, only an extreme dedication to free/libre and bootstrapping from a minimal binary
09:03:34 <gjanssens> And for what it's worth flatpaks use similar build scripts
09:04:17 <gjanssens> They are all variations of defining a way to set up a consistent environment.
09:04:24 <chris> if it's similar then are there instructions to set up a dev environment using docker/flatpak? are these scripts merged already?
09:04:42 <chris> 'would be nice' to document how to run it somehwere
09:05:52 <gjanssens> Well, it's not exactly implemented like that, nor documented.
09:07:33 <chris> would docker allow you to set up all dependencies in the environment and start hacking ?
09:07:48 <chris> ^ thence the appeal of guix environment
09:08:28 <gjanssens> It certainly can, but I don't know the exact details of what guix does
09:08:30 <chris> (btw the MAKE_LINKS trick works well and is instrumental in modularising (reports))
09:08:57 <chris> ^ guix I'll report back if I manage to make it work :)
09:09:38 <gjanssens> You'd create a docker image using a docker file similar to what we use for ci testing
09:09:59 <gjanssens> And then you run a container based on that image, enter the container and off you go
09:10:41 <chris> ^ ok - all it needs then is a wiki section
09:11:09 <gjanssens> And while thinking about this, fedora offers a "toolbox" feature in recent releases.
09:11:43 <gjanssens> A toolbox is kind of an environment shielded from the main system.
09:11:55 <gjanssens> You could script that as well to set up a build environemnt for gnucash.
09:12:13 <gjanssens> I haven't played with it but it's again targeting the same use case.
09:16:25 *** sbluhm has quit IRC
09:22:43 *** giuseppef has joined #gnucash
09:22:44 *** ChanServ sets mode: +v giuseppef
09:25:14 *** pim has left #gnucash
09:25:43 *** giuseppef has quit IRC
09:30:36 <chris> !@#$ my qif-objects modularisation error was caused by an errant parens!
09:45:08 *** User_ has quit IRC
09:45:37 *** sbluhm has joined #gnucash
09:45:37 *** ChanServ sets mode: +v sbluhm
10:09:29 *** TommyT has joined #gnucash
10:09:29 *** ChanServ sets mode: +v TommyT
10:14:55 <TommyT> I recently experienced data loss / corruption and am unsure how to proceed. It disturbs me because I found two different problems -- 6 or 8 transactions that were marked uncleared and at least one completely missing despite several successful statement reconciliations.
10:15:51 <TommyT> It bothers me because I THOUGHT that should be impossible -- I thought the reconciliation was based on doing the math all the way back to the beginning
10:16:45 <TommyT> I have used GnuCash for quite a few years at this point.
10:17:15 *** David has joined #gnucash
10:18:44 <TommyT> Around the time of the missing / corrupted transactions I switched from a self compiled 3.8 to flatpak 4.x
10:19:48 <TommyT> These transactions were from September and earlier and as I said I've had several successful account reconciliations since. I found them because I was looking for why the checks had not cleared, but there they were in the bank statement
10:20:20 <warlord> TommyT, did you, perhaps, open a backup file by accident?
10:20:49 <TommyT> I will look for that, though if that is true how could I have done a couple of months' reconciliations since?
10:21:38 <TommyT> I made a backup of the data file before switching to flatpak and if I recall I intentionally installed 3.9 4.0 etc to open the data file just in case
10:21:59 <TommyT> to let it do whatever it wanted to the data file
10:22:11 <TommyT> I also ran the repair
10:23:13 <warlord> well, if you opened a backup by accident, it could have forked your data; so if you reconcile on Dec 3, but then go and open Dec 1, you may have lost data (including the reconciliation).
10:23:35 <warlord> Another place to look: could you have accidentally modified a transaction that caused it to un-reconcile?
10:23:41 <warlord> Are your current balances correct?
10:25:13 <TommyT> I had completed a successful reconcile to the last statement 11/15; that's all I know about the balances. But it turns out despite the successful reconcile I was missing at least one transaction from June or so and a half dozen checks were marked uncleared from Septemberish
10:25:56 <TommyT> I never double click my gnucash files to launch. In fact this one is so old it doesnt end in .gnucash
10:26:09 <warlord> Were you missing it on 11/15?
10:26:10 <TommyT> I use launchers to always launch the correct file, but maybe I moved something
10:26:21 *** Mechtilde has quit IRC
10:26:46 <warlord> Or maybe you clicked on a backup file. What does GnuCash say the filename is?
10:27:19 <TommyT> I'm not at the computer but the filename is something like TnA Financials (no extension)
10:27:49 <TommyT> As I said I never double click GnuCash -- running Ubuntu and I created GNOME launchers for all my data files (I have about nine active books)
10:28:02 <TommyT> This is my personal books with the oldest data going back years
10:28:39 <warlord> Ah, okay, that's what you mean by launcher, not "double click on file"
10:28:45 <TommyT> Yes that
10:29:10 <TommyT> I have a custom launcher with menu items for all the files I regularly open
10:29:16 <warlord> ok
10:29:34 <warlord> what file storage mechanism do you use? XML or SQL?
10:29:41 <TommyT> XML
10:29:54 <TommyT> compressed; "normal" GnuCash data file
10:32:09 <TommyT> As I mentioned, around this time I upgraded GnuCash through several revisions, though I was careful to install interim FlatPak versions just in case
10:33:20 <TommyT> I make data backups monthly so I can go to the bank and retrieve the backups back through to July or so and I might be able to pinpoint plus or minus 15 days when things went missing,
10:33:38 <TommyT> HOWEVER I worry that this is just one account in a very complicated GnuCash book
10:33:43 <TommyT> so there may be other things missing
10:35:09 <warlord> One thing you can try: decompress some data files and run a diff on the XML to see what differences there are.
10:35:21 <warlord> They SHOULD be well-ordered these days to make that relatively easy.
10:35:51 <TommyT> I thought I could trust statement reconciliation. How is it possible to reconcile with missing and uncleared transactions?
10:36:32 <TommyT> This data file will be very large, I think
10:40:28 <TommyT> 3.1 MB doesn't sound so large but I know it has lots of transactions
10:45:36 <warlord> Well, the diff should be relatively limited, and will show you where to look.
10:47:07 <warlord> Reconciling is based on the balance. So if you have a starting balance of $1234 and an ending balance of $2345, you will need to mark off a net change of $1111. HOW you get that $1111 doesn't matter. So if you have multiple errant transactions that add up to $1111, then you could reconcile without reconciling all transactions.
10:47:46 <warlord> So if you successfully reconciled on 11/15, then that means the ending balance (and marked transactions) on 11/15 matched whatever you entered into the reconcile dialog.
10:48:25 <warlord> If there are missing transactions, then either they were removed/modified AFTER 11/15.. Or they are errant, so not included.
10:48:41 <warlord> Is the current reconciled balance of the account correct?
10:48:43 <TommyT> When I was trying to fix it, I made a copy of the data file and went back and un-cleared to June, and kept finding more things missing
10:48:57 <TommyT> I'm digging up data files now
10:49:30 <TommyT> Yes I successfully reconciled for several months with the missing / uncleared transactions
10:50:34 <warlord> Then that means those transactions are not actually missing or valid, otherwise the reconciliations would have failed.
10:50:52 <warlord> ... or you entered the wrong reconcile ending balance.
10:51:00 <TommyT> That's the weird part, isn't it?
10:52:04 <TommyT> I just uncompressed the data files and the backup I made in august or september (or october?) the timestamps are weird is 34.8MB and the most recent is 25.6MB
10:53:03 <TommyT> I think I will need to retrieve some backup images from the bank to be sure of the dates
10:54:47 <chris> gjanssens: you could help me modularise on 1 issue: how to import gnc-html into guile? what module name does it use?
10:55:09 <chris> it's needed for gnc-build-url and URL-TYPE-REGISTER
10:55:57 <warlord> TommyT, very weird that a latter file would be smaller than a former file.. did you have a disc failure?
10:56:47 <TommyT> Not that I know of; no other errors that I have seen... No errors uncompressing.
10:57:35 *** ArtGravity has joined #gnucash
10:57:35 *** ChanServ sets mode: +v ArtGravity
10:58:41 <TommyT> I was apparently sloppy when I created my backup directory (renamed it august but the contents are September) and apparently I intentionally opened the file in October but the gnucash backup files are in the directory
10:59:01 <TommyT> so I have a lot of data to sift through
11:00:12 <TommyT> In other words I have a snapshot of activity back in September from a complete 30 day gnucash automatic backup directory but I confused myself by renaming it funny
11:01:43 <TommyT> Actually I figured out it IS october; I have my directory sorted backwards. Maybe I need to stop and make some more backup copies before I completely confuse myself
11:02:23 <warlord> Take your time.
11:14:02 <chris> gjanssens: from your understanding, does docker containerisation allow packages to be cached? ie if docker defines 'apt update' does it always download? because guix will cache them
11:15:26 <gjanssens> Docker does cache. If you don't make changes to the dockerfile and don't remove the generated image, it will be reused.
11:17:37 <gjanssens> chris: there are a few modules related to gnc-html, I suspect you're after (gnucash html) ? That's in gnucash/html/gnucash-html.scm
11:18:02 <gjanssens> Sorry in gnucash/html/html.scm
11:18:08 <chris> right, missed it
11:18:10 *** bertbob has quit IRC
11:18:34 <chris> otherwise I think modularisation is 'nearly done'
11:19:00 <gjanssens> np and nice. I don't know if I can review though :(
11:19:40 <chris> np the build log will review for you
11:20:52 *** sbluhm has quit IRC
11:25:32 <TommyT> Warlord: comparing the files the first diff is what I assume is the total transaction count. In October it shows as 20598 and in December it shows as 20812 so a couple hundred transactions might be right. The rest of the diff is showing transactions but I'm not sure what I'm looking for.
11:25:48 *** chris has quit IRC
11:26:03 *** chris has joined #gnucash
11:26:03 *** gncbot sets mode: +o chris
11:26:03 *** ChanServ sets mode: +v chris
11:26:07 <TommyT> I am trying to decide whether it's best to go to my October 2 backup and try upgrading again
11:26:08 *** bertbob has joined #gnucash
11:26:08 *** ChanServ sets mode: +v bertbob
11:26:25 <TommyT> upgrading meaning opening the file in a newer GnuCash
11:26:36 *** jcarl43 has joined #gnucash
11:26:36 *** ChanServ sets mode: +v jcarl43
11:26:42 <TommyT> and then entering the couple hundred transactions since October
11:27:53 <fell> chris, commit 9e23e41 is the contrary of modularizarion.
11:32:55 <warlord> You shouldn't have to enter the new transactions...
11:32:59 <warlord> Do you still have the log files?
11:33:28 <TommyT> warlord: as an experiment I just renamed my two uncompressed data files and opened them in current release (Flatpak) GnuCash. The October file opens fine. The December file was taking a long time and I killed GnuCash leaving a lock.
11:33:45 <TommyT> I'm going to experiment some more
11:35:32 <warlord> Define "long time"?
11:35:45 <warlord> Yes, it will leave a lock if you kill the process.
11:37:51 <TommyT> It took more than a minute; the October one (pre conversion) opened instantly
11:39:36 <TommyT> ACTUALLY what happens is the file SAYS it's open but the Accounts panel is empty. The progress bar has finished and the application looks open except there's nothing to see
11:40:00 <TommyT> The file dirty asterisk is set at the top of the window.
11:40:27 <TommyT> But when I quit GnuCash it doesn't delete the lock
11:40:40 <TommyT> Despite me closing it properly
11:40:43 <TommyT> VERY odd
11:41:03 <warlord> TommyT, maybe there is a .gcm file and it's missing the CoA page?
11:41:11 <warlord> Did you try to open a new CoA page?
11:41:41 <TommyT> I will try again; I don't know about a .gcm file I renamed the xml file .gnucash
11:42:06 <TommyT> It created a .LNK and a .LCK file but did not delete them when I quit without saving
11:42:13 *** jervin has joined #gnucash
11:43:01 <TommyT> I will open the uncompressed XML file I've named TnA_Financials (1204).gnucash
11:43:05 *** guak has joined #gnucash
11:43:52 <warlord> gcm is the Metadata file -- stored in your .gnucash storage.
11:43:59 <TommyT> After a slow-ish file open (with the progress bar) I get the window I mentioned with the dirty file asterisk and an Accounts tab but nothing below it
11:44:33 *** sbluhm has joined #gnucash
11:44:33 *** ChanServ sets mode: +v sbluhm
11:44:42 <warlord> TommyT, View -> New Accounts Page
11:44:46 <TommyT> I should say the window has its decoratioins -- $grand total, Net Assets Account Name Description, but no ACCOUNTS listed
11:45:17 <TommyT> Now I have TWO empty accounts tabs
11:45:25 <TommyT> after New Accounts Page
11:45:27 <warlord> Ummmmm......
11:45:57 <TommyT> I think this definitively identifies the data file as borked
11:46:09 <TommyT> but borked in an odd and interesting way
11:46:13 <warlord> No, not necessarily.
11:46:24 <warlord> Anything in the trace file?
11:46:32 *** jervin has quit IRC
11:46:45 <TommyT> I'll have to remember how to find a trace file in a flatpak GnuCash
11:48:30 *** frakturfreak has quit IRC
11:51:41 <TommyT> flatpak run org.gnucash.GnuCash --logto stdout
11:51:41 <TommyT> Found Finance::Quote version 1.49.
11:51:41 <TommyT> * 10:50:38 ERROR <gnc.backend.file.sixtp> parse failed at:
11:51:41 <TommyT> (stack-frame 0x563752f74d80
11:51:41 <TommyT> (line -1) (col -1)
11:51:42 <TommyT> (parser 0x563752f6f200)
11:51:44 <TommyT> (tag (null))
11:51:46 <TommyT> (data-for-children (nil))
11:51:48 <TommyT> (data-from-children)
11:51:52 <TommyT> (frame-data (nil)))
11:51:54 <TommyT> (stack-frame 0x563752f76c70
11:51:56 <TommyT> (line 31) (col 54)
11:51:58 <TommyT> (parser 0x563752f6f260)
11:52:00 <TommyT> (tag gnc-v2)
11:52:02 <TommyT> (data-for-children (nil))
11:52:04 <TommyT> (data-from-children ((tag gnc:count-data) (data 0x563752f4e000)))
11:52:06 <TommyT> (frame-data (nil)))
11:52:08 <TommyT> (stack-frame 0x563752f76ed0
11:52:10 <TommyT> (line 33) (col 26)
11:52:12 <TommyT> (parser 0x563752f6f2c0)
11:52:14 <TommyT> (tag gnc:book)
11:52:16 <TommyT> (data-for-children (nil))
11:52:18 <TommyT> (data-from-children ((tag gnc:transaction) (data 0x563754b192b0)) ((tag gnc:transaction) (data 0x563754b00d20)) ((tag gnc:transaction) (data 0x563754b0be70)) ((tag gnc:transaction) (data 0x563754adbed0)) ((tag gnc:transaction) (data 0x563754b123c0)) ((tag gnc:transaction) (data 0x563754a9d330)) ((tag gnc:transaction) (data 0x563754ae8540)) ((tag gnc:transaction) (data 0x56375493f840)) ((tag gnc:transaction) (data 0x563754b10b60)) ((tag
11:52:25 <TommyT> gnc:transaction) (data 0x563754ae5900)) ((tag gnc:transaction) (data 0x563754afafb0)) ((tag gnc:transaction) (data 0x563754ae9a60)) ((tag gnc:transaction) (data 0x563754aaf620)) ((tag gnc:transaction) (data 0x56375493f840)) ((tag gnc:transaction) (data 0x563754a6edd0)) ((tag gnc:transaction) (data 0x563754ae8700)) ((tag gnc:transaction) (data 0x563754adb840)) ((tag gnc:transaction) (data 0x563754b0f200))
11:52:30 <TommyT> goes on for awhile
11:52:47 <TommyT> (data 0x563752f4e000)))
11:52:47 <TommyT> (frame-data (nil)))
11:52:47 <TommyT> (stack-frame 0x563754b19da0
11:52:47 <TommyT> (line 698386) (col 33)
11:52:47 <TommyT> (parser 0x563752f6fde0)
11:52:47 <TommyT> (tag gnc:transaction)
11:52:49 <TommyT> (data-for-children 0x563754b10a10)
11:52:53 <TommyT> (data-from-children)
11:52:55 <TommyT> (frame-data 0x563754b10a10))
11:52:57 <TommyT> (stack-frame 0x563754b19de0
11:52:59 <TommyT> (line 698387) (col 22)
11:53:01 <TommyT> (parser 0x563752f6fde0)
11:53:03 <TommyT> (tag trn:id)
11:53:05 <TommyT> (data-for-children 0x563754ad1320)
11:53:07 <TommyT> (data-from-children)
11:53:09 <TommyT> (frame-data (nil)))
11:53:11 <TommyT> * 10:50:38 WARN <gnc.backend> [GncXmlBackend::load()] Syntax error in Xml File /home/twt/Documents/GnuCash repair/TnA_Financials (1204).gnucash
11:53:14 <TommyT> The above is from the beginning and end of the console output
11:54:31 <warlord> Looks like there is an actual error in the data file. Check out line 698387. Looks like a bad transaction id?
11:55:33 <TommyT> that is the last line in the file:
11:55:34 <TommyT> <trn:id type="guid">f0587f999f15
11:55:54 <TommyT> so maybe the file got truncated
11:56:00 <chris> fell: agreed I've reverted in the PR
11:56:29 <warlord> indeed, it did.
11:57:33 <TommyT> yeah pretty obviously truncated... but at this moment I don't know where or when the truncation happened
11:57:51 <TommyT> Since I'm working with an uncompressed copy
11:58:30 <TommyT> I will start looking back through the most recent GnuCash backups but I know the uncleared checks have been plaguing me for at least a few weeks
11:59:26 <TommyT> so again I may decide to check the SMART status on the hard drive (I have a replacement in hand just in case) and look for other missing things, as best I can
11:59:56 <warlord> I would start by looking at the file sizes and find where the file-size goes down... even in compressed form.
12:00:15 <TommyT> That's the weird part -- compressed they look pretty similar
12:00:37 <TommyT> I know years and years ago I "rescued" a file that got a spurious control Z in it
12:00:58 <TommyT> so maybe somwhere a cosmic ray flipped a bit
12:01:02 *** sbluhm has quit IRC
12:02:09 <warlord> Could be.. BUT if that were the case, it would have thrown errors from there on..
12:03:32 <warlord> Or the decompress failed... Disk full?
12:03:44 <warlord> (or interrupted?)
12:05:03 *** storyjesse has quit IRC
12:05:12 <TommyT> Home shows 28% full
12:05:28 <warlord> what about /tmp? Or other partitions?
12:06:11 <TommyT> Only the silly "snap" partitions (this is an Ubuntu 18.04 box)
12:06:22 <TommyT> everything else is fine
12:07:00 <TommyT> I haven't looked at SMART status in forever so I have to remember how to find it
12:07:08 <warlord> Can you try (again) to decompress the data file and see if it does the same thing?
12:07:33 <TommyT> OK
12:08:21 <TommyT> That's actually a good thought because I'm on a SSH connection to the box with the original data; working on a different machine
12:12:44 <TommyT> OK so the earlier failure was a false alarm -- the fresh copy and decompress works fine
12:12:57 <warlord> okay. that's good.
12:13:12 <warlord> so back to this missing data...
12:13:46 <warlord> I would open up the latest data file, open the account in question and look at the reconciled balance. Is it correct?
12:15:15 <TommyT> I have a figure but I'll have to pull up the statement
12:16:39 <warlord> if it's correct, then any "missing" or "unreconciled" transactions are invalid.
12:16:49 <warlord> or mis-filed
12:17:05 <TommyT> The numbers differ, but that's not necessarily a problem because I use a subaccount to keep from incurring a minimum balance fee
12:17:13 <TommyT> I'm looking some more
12:18:00 <TommyT> OH look at that -- the "missing" transaction accidentally got into the subaccount
12:18:09 <warlord> Well, there you go!
12:18:13 <TommyT> It's reconciled
12:18:23 <TommyT> That does not explain all the unreconciled transactions
12:19:07 <warlord> Either they are duplicates, misfiled (belong to a different account), or they never got processed.
12:20:01 <TommyT> See THAT'S the weird part -- they're all to a particular healthcare provider and I was confused that some old checks had never been cashed
12:20:16 <TommyT> They're all in this account, and no duplicates
12:20:35 <TommyT> I have had a report open for months with them in it but I don't see how that would affect them
12:21:06 <TommyT> It's also possible I had a search window open with them all in them at some point but I certainly would not have gone through unreconciling them
12:21:17 <warlord> Well, if you have a check that was never cashed, that would explain an unreconciled transaction.
12:22:22 <TommyT> What I mean is they WERE cashed and apparently reconciled, but somehow they got unreconciled somehow
12:22:41 <TommyT> and I was able to reconcile multiple month statements afterwards which makes no sense
12:22:49 <warlord> Then how would your reconciled balance be correct/?
12:22:57 <TommyT> I have no idea
12:22:59 <warlord> Are you sure they were not entered twice?
12:23:20 <warlord> .. and you reconciled one
12:23:42 <TommyT> Looking for that now
12:24:28 <warlord> View -> Filter by ... Disable reconciled. Then find one that should be reconciled, and "jump" to the other side?
12:24:48 <warlord> Or then replace the filter and look at other transactions nearby?
12:24:57 <warlord> Or search for transactions with the same amount?
12:26:05 <TommyT> I just searched for transactions to the provider and yes I'm finding duplicates. Probably I re-entered a batch in a half awake stupor
12:26:26 <warlord> And there you go :)
12:26:39 <warlord> (and this is why we *have* reconciliation) :)
12:26:57 <warlord> If the reconcile finishes, then anything left over is, most likely, an error.
12:27:16 <TommyT> Yes! Exactly. The "missing" transaction wasn't missing, just tucked in the wrong subaccount
12:27:35 <TommyT> Thank you for your help. I'm very glad I'm not going to have to re-enter data from months ago
12:27:38 <warlord> Moving it would probably invalidate the reconciliation
12:27:44 <warlord> Me too!!! :)
12:28:27 <TommyT> Even if it does invalidate the reconciliation (moving the transaction from the subaccount to the main account) I can easily see it when I reconcile next
12:28:48 <warlord> yes
12:30:41 <TommyT> I've been using GnuCash for a long long time and this is the first time I've had this sort of problem; usually I can figure it out. I actually recently found a mis-entered transaction from the early 2000s because I decided to go back and put in more data; at some point I couldn't find it and entered a correcting transaction but for completeness I entered an UN-correcting transaction to fix it. My excuse then was I had a serious medical condition and
12:30:41 <TommyT> probably shouldn't have been doing my bookkeeping
12:31:14 <TommyT> Now it's probably just sleep deprivation
12:34:20 <warlord> I just reconciled some accounts back to 2013 and found some data entry issues.
12:39:42 *** giuseppef has joined #gnucash
12:39:43 *** ChanServ sets mode: +v giuseppef
12:48:42 *** giuseppef has quit IRC
12:49:05 *** giuseppef has joined #gnucash
12:49:06 *** ChanServ sets mode: +v giuseppef
12:54:12 *** Mechtilde has joined #gnucash
13:10:44 *** sbluhm has joined #gnucash
13:10:44 *** ChanServ sets mode: +v sbluhm
13:36:52 *** hfxpaul has joined #gnucash
13:36:52 *** halipaul has quit IRC
13:48:12 *** Mechtilde has quit IRC
14:02:35 *** Mechtilde has joined #gnucash
14:05:28 <jralls> warlord, I've finished editing bug-writing.html.tmpl. I propose to create a new git repo gnucash-bugzilla with directory templates/en/custom to hold it and any other templates we decide to customize.
14:08:56 <warlord> jralls, okay. We'll need to figure out how to post-hook copy into BZ
14:10:49 <jralls> I'd think that a straight cp would be easiest.
14:12:03 *** David has quit IRC
14:12:08 *** David has joined #gnucash
14:26:42 <warlord> Except that it's a bare repo in gitolite.
14:27:28 <warlord> So I need to have a checkout elsewhere with a post-commit-hook script that calls git pull + cp
14:32:24 <jralls> Or turn BZ
14:32:34 <jralls> 's dir into a working dir.
14:35:00 <warlord> Sure, I could try to make /usr/share/bugzilla/template/en/custom as a git checkout.
14:35:27 <jralls> The the hook in the gitolite repo is #!/bin/sh; /usr/bin/newgrp xxx; GIT_WORK_TREE=/path/to/BZ
14:35:53 <jralls> xxx is the group that owns the BZ dir.
14:36:38 <jralls> I use that on my local server to work the devel version of a website.
14:41:45 <jralls> The WD doesn't even need a .git/
14:51:55 <warlord> The tree needs to live in /usr/share/bugzilla/template/en/custom -- and needs to match /usr/share/bugzilla/template/en/default (for anything that we override).
14:52:21 <warlord> If you think you can do that without a git checkout, great. But I'm not so sure.
14:53:23 <jralls> What user:group owns /usr/share/bugzilla/template/en/custom?
14:53:29 <warlord> The tree is owned root:apache (vs git:gitolite3)
14:53:48 <warlord> (note that 'custom' does not yet exist, but I can mkdir it)
14:58:25 <jralls> OK. You'll also need to add user git to group apache and set the custom/'s privs to 775.
15:11:49 <warlord> I need to think if I actually like that idea..
15:12:13 <warlord> It would mean anyone who can break gitolite access could do anything to the web service.
15:16:20 <jralls> Hmm. I don't remember: Will apache barf if that directory belongs to apache:other_group?
15:16:53 *** gjanssens has quit IRC
15:18:10 <warlord> I don't know, but it does tell me I'd rather have a triggered pull
15:19:26 <jralls> Otherwise we could point the WD somewhere else and have a shell script that belongs to root:apache with `cp .* /usr/share/bugzilla/template/en/custom` and fire that off in the hook.
15:21:22 *** frakturfreak has joined #gnucash
15:21:22 *** ChanServ sets mode: +v frakturfreak
15:21:33 <jralls> Or I guess git init .../en/custom and have a shell script that does git pull from there that's run by the hook.
15:22:31 <jralls> BTW, what version of gitolite is on code? Apparently we can't have per-repo hooks unless it's 3.6 or later.
15:23:01 <jralls> https://gitolite.com/gitolite/cookbook#v36-variation-repo-specific-hooks
15:25:16 <warlord> gitolite3-3.6.11-1.fc29.noarch
15:26:43 <jralls> Good. There are a couple of things you need to check in .gitolite-rc or whatever it's called. That's at the beginning of that link.
15:30:34 <warlord> jralls, okay, #1 was already done (although it's "local-code", not "local"). But #2 appears to be more of a challenge.
15:31:31 <jralls> I figured the first one was in place. There's no ENABLE list?
15:32:10 <warlord> nope
15:32:11 *** giuseppef has quit IRC
15:32:20 <warlord> just PRE_CREATE and POST_CREATE lists
15:35:39 <jralls> Here's some instructions for upgrading to a 3.6 rc file: https://gitolite.com/gitolite/rc.html#appendix-a-upgrading-the-rc-file
15:36:27 <warlord> jralls, we can just update the existing post-receive script .
15:36:46 <warlord> We already do repo-specific things in there.
15:38:32 <warlord> I'll add upgrading the rc file to my to-do list.
15:39:47 <jralls> Ah, true. if [ "$reponame" = "gnucash-bugzilla" ]; then /usr/local/bin/pull-bz-template-update; fi (or whatever you want to call the pull shell-script).
15:43:38 *** giuseppef has joined #gnucash
15:43:38 *** ChanServ sets mode: +v giuseppef
15:43:56 *** giuseppef has quit IRC
15:44:39 *** giuseppef has joined #gnucash
15:44:39 *** ChanServ sets mode: +v giuseppef
15:52:41 <warlord> Yep
16:53:02 *** bertbob has quit IRC
16:59:39 *** delphijustin has joined #gnucash
17:02:17 *** bertbob has joined #gnucash
17:02:17 *** ChanServ sets mode: +v bertbob
17:03:19 <delphijustin> does gnucash allow getting transactions from banks automatically like Quicken?
17:04:43 *** sbluhm has quit IRC
17:05:19 <jralls> delphijustin depends on your bank. If you're in Australia, France, or the US and it supports OFX Direct Connect, or you're Germany, then yes. Otherwise probably not.
17:05:22 *** KaiForce has quit IRC
17:05:37 <jralls> *in* Germany.
17:06:29 <delphijustin> I'm using ally in USA
17:07:12 *** bertbob has quit IRC
17:09:16 <jralls> Then it depends on whether they support OFX Direct Connect. See if they're in http://ofxhome.com's database and if they've had a successful validation recently.
17:10:45 *** bertbob has joined #gnucash
17:10:46 *** ChanServ sets mode: +v bertbob
17:28:00 <chris> good progress in modularisation. now I think all I need to do is audit qif C code to ensure the scm code is exported
17:28:08 <fell> jralls aqbaning supports also EBICS in CH, DE and FR. And PayPal woldwide.
17:29:05 <jralls> I thought EBICS was obsolete and that France used OFX.
17:30:22 <fell> https://en.wikipedia.org/wiki/Electronic_Banking_Internet_Communication_Standard
17:31:51 <fell> or official https://www.ebics.org/en/home
17:32:29 *** delphijustin has left #gnucash
17:33:31 <fell> July 2020: Austria joins EBICS SC
17:33:54 <geneAum> in a Stock account, the Action column has Buy, Dist, Dividend, Fee, Income, Interest, LTCG, Price, Sell, Split, STCG. What impact does it have to set/use this field in the split? Do any reports use that field? Does anything else?
17:34:45 <fell> You can overwrite the defaults.
17:35:26 <fell> Without looking in the code, I assume you can use it to filter.
17:36:51 <jralls> geneAum: You can use it to find or filter. Some reports include it. it's free text, you can put anything you want in it, and you can optionally use it for the num field in the transaction line.
17:38:19 <jralls> chris is working on a feature that makes more direct use of it in a stock transaction entry widget, but that won't be ready for a while.
17:38:20 *** giuseppef has quit IRC
17:38:51 *** giuseppef has joined #gnucash
17:38:51 *** ChanServ sets mode: +v giuseppef
17:41:14 <chris> jralls: it's stalled until I get a hint ;)
17:46:28 <jralls> chris, it's stalled until you redesign it bass-ackwards decomposition.
17:46:39 <jralls> s/it/its/
17:47:42 <chris> i wont deny; it was meant as a prototype first. not sure whats the current best pracrice
17:52:20 <jralls> Has nothing to do with current best practice. Has to do with logically mapping 12 investment actions to a small number of functions.
17:53:40 <chris> well, it's not yet frozen because I'd wish CDB to start poking at it when he can and this wont happen unless there's a win32 build
17:54:42 <jralls> There's no point in building anything until you fix the design.
17:54:56 <chris> but how??
17:56:31 *** Herbert has joined #gnucash
17:56:51 *** giuseppef has quit IRC
17:57:02 *** Herbert has quit IRC
17:58:06 <jralls> I told you that a month ago. Use an enum on the 12 actions (buy, sell, div, notional, split, buy back, sell short, short div, short notional, short split, open long, open short). Write functions for each of the half-dozen split types you need to create. Use a switch statement on the enum to call the split-creation functions needed for that enum value.
17:59:24 <jralls> I think you've already got the set of functions, you just have a convoluted way of selecting which ones to call.
18:15:11 <CDB-PHONE_> I agree with use of enumerator and switch statement
18:16:31 <jralls> CDB-PHONE_ I didn't know you knew what that meant!
18:16:57 <CDB-PHONE_> I last coded Java 1.4.1 back in 2008
18:17:01 <CDB-PHONE_> 2009*
18:17:10 <CDB-PHONE_> so I'm not COMPLETELY clueless
18:17:22 <CDB-PHONE_> just extremely rusty
18:19:22 <jralls> Nah, it's like riding a bike. It'll come right back, albeit with a few scrapes and bruises. ;-)
18:41:37 *** bertbob has quit IRC
18:46:40 *** bertbob has joined #gnucash
18:46:40 *** ChanServ sets mode: +v bertbob
18:49:53 <chris> ooh CDB knows his malloc and free
18:53:18 <CDB-PHONE_> what's Malloc and free?
18:53:40 <chris> or maybe not, java does gc
19:11:50 <CDB-PHONE_> yeah Java garbage collection and memory allocation
19:28:35 *** chris has quit IRC
19:29:28 *** chris has joined #gnucash
19:29:28 *** gncbot sets mode: +o chris
19:29:28 *** ChanServ sets mode: +v chris
19:59:03 *** jcarl43 has quit IRC
20:34:56 <chris> jralls if you're skilled with docker, it "would be nice" to have wiki instructions how to set up a dev environment by running a container
20:52:06 *** geneAum has quit IRC
20:56:12 *** Mechtilde has quit IRC
21:00:24 *** Mechtilde has joined #gnucash
21:07:30 *** bertbob has quit IRC
21:09:06 *** bertbob has joined #gnucash
21:09:06 *** ChanServ sets mode: +v bertbob
21:10:36 *** Mechtilde has quit IRC
21:24:48 *** bertbob has quit IRC
21:26:29 *** bertbob has joined #gnucash
21:26:29 *** ChanServ sets mode: +v bertbob
21:31:13 <warlord> I'm going to upgrade my laptop.. I'll be back online some time tomorrow.
21:36:57 *** warlord has quit IRC
22:07:02 *** guak has quit IRC
23:01:57 *** TommyT has quit IRC
23:18:38 *** ArtGravity has quit IRC
23:44:20 *** jervin has joined #gnucash
23:54:00 *** Mechtilde has joined #gnucash
23:56:30 *** David has quit IRC
23:56:36 *** David has joined #gnucash