2020-12-03 GnuCash IRC logs
00:08:55 *** angel has quit IRC
00:17:20 *** o01eg has quit IRC
00:30:24 *** CDB-PHONE_ has joined #gnucash
00:30:28 *** CDB-PHONE_ has joined #gnucash
00:31:26 *** CDB-PHONE_ has quit IRC
00:31:27 *** CDB-PHONE has quit IRC
00:31:29 *** CDB-PHONE_ has joined #gnucash
00:31:48 *** CDB-PHONE_ has joined #gnucash
01:18:42 *** fell has quit IRC
01:19:02 *** fell has joined #gnucash
01:19:02 *** ChanServ sets mode: +o fell
01:41:57 *** Mechtilde has joined #gnucash
01:43:07 *** sbluhm has joined #gnucash
01:43:07 *** ChanServ sets mode: +v sbluhm
01:59:33 *** fell has quit IRC
02:00:24 *** fell has joined #gnucash
02:00:25 *** ChanServ sets mode: +o fell
02:17:33 *** giuseppef has quit IRC
02:17:37 *** giuseppef has joined #gnucash
02:17:37 *** ChanServ sets mode: +v giuseppef
02:40:09 *** Mechtilde has quit IRC
03:22:10 *** storyjesse has quit IRC
03:40:15 *** bertbob has quit IRC
03:52:16 *** giuseppef has quit IRC
03:52:36 *** giuseppef has joined #gnucash
03:52:36 *** ChanServ sets mode: +v giuseppef
03:54:31 *** bertbob has joined #gnucash
03:54:32 *** ChanServ sets mode: +v bertbob
04:05:19 *** gjanssens has joined #gnucash
04:05:19 *** ChanServ sets mode: +o gjanssens
04:06:44 *** hussam has quit IRC
04:08:12 *** Mechtilde has joined #gnucash
04:08:12 *** Aussie_matt has quit IRC
05:07:53 *** Mechtilde has quit IRC
05:09:52 *** PowaBanga has joined #gnucash
05:13:50 *** storyjesse has joined #gnucash
05:14:34 *** Mechtilde has joined #gnucash
05:24:05 *** User_ has joined #gnucash
05:34:11 *** User_ has quit IRC
05:50:31 *** Herbert has joined #gnucash
06:02:49 *** Aussie_matt has joined #gnucash
06:10:48 *** hussam has joined #gnucash
06:10:48 *** ChanServ sets mode: +v hussam
06:20:18 *** Mechtilde has quit IRC
06:55:53 *** Mechtilde has joined #gnucash
06:57:34 *** Aussie_matt has quit IRC
07:25:18 <chris> gjanssens: funny. within 1 set() group, there are many scm files. They're tagged MAKE_LINKS. One of them didn't see a later one. I swap the order and now the later sees the earlier one.
07:25:45 <chris> I trust this is just a glitch
07:50:52 *** KaiForce has joined #gnucash
07:54:19 <gjanssens> Probably indeed. Thanks for fixing
07:54:19 <gncbot> gjanssens: Sent 13 hours and 28 minutes ago: <chris> other public modules tend to use define + #:export rather than define-public
07:54:20 <gncbot> gjanssens: Sent 13 hours and 22 minutes ago: <chris> so, (engine c-interface) must includes same headers as (engine), and can also depend on (engine business-core). obvious in hindsight.
07:54:51 <gjanssens> Right
08:01:59 <geneAum> i made a Stock account that has a Liability account as its parent, in order to track borrowed stock. is that not a crazy idea? one consequence is that the Balance Sheet lists this Stock Liability under Assets and not under Liabilities. is that a bug, or is this just a foolish use case?
08:07:30 *** frakturfreak has joined #gnucash
08:07:30 *** ChanServ sets mode: +v frakturfreak
08:09:55 *** chris-phone has joined #gnucash
08:09:55 *** ChanServ sets mode: +v chris-phone
08:10:32 <chris-phone> It would be more clever to track the borrowed stock as negative stock balance, but under an asset account
08:11:00 <chris-phone> I don't know whether the reports can handle negative stock balance. Are you short selling?
08:24:28 *** rudi has joined #gnucash
08:26:44 *** hussam has quit IRC
08:28:56 *** chris-phone has quit IRC
08:29:28 *** fell has quit IRC
08:29:47 *** fell has joined #gnucash
08:29:47 *** ChanServ sets mode: +o fell
08:33:47 *** Jimraehl1 has joined #gnucash
08:34:38 *** Jimraehl1 has quit IRC
08:42:05 *** mrBussy has joined #gnucash
08:43:37 *** mrBussy has quit IRC
09:08:16 *** hussam has joined #gnucash
09:08:16 *** ChanServ sets mode: +v hussam
09:13:35 <chris> more progress :)
09:15:54 <warlord> yay
09:19:44 <chris> more progress = now tacking the individual reports
09:35:28 <chris> the modularisaiton will have piecemeal commits: https://github.com/christopherlam/gnucash/tree/maint-modularisation
09:35:32 <warlord> refactor? ;)
09:36:19 <chris> warlord: no it means I'm nearly done dependencies, but not yet fixed individual reports' imports
09:36:52 <warlord> ah
09:36:58 <warlord> ok
10:10:43 *** fell has quit IRC
10:11:02 *** fell has joined #gnucash
10:11:03 *** ChanServ sets mode: +o fell
10:23:30 *** fell has quit IRC
10:29:16 *** fell has joined #gnucash
10:29:16 *** ChanServ sets mode: +o fell
10:30:19 *** Mechtilde has quit IRC
10:34:45 *** fell has quit IRC
10:35:46 *** jervin has quit IRC
10:37:57 *** fell_laptop has joined #gnucash
10:37:57 *** ChanServ sets mode: +o fell_laptop
10:47:43 *** lmat has quit IRC
10:50:37 *** sbluhm has quit IRC
10:53:56 *** o01eg has joined #gnucash
11:01:41 *** lmat has joined #gnucash
11:02:59 *** storyjesse has quit IRC
11:09:26 *** sbluhm has joined #gnucash
11:09:26 *** ChanServ sets mode: +v sbluhm
11:11:40 *** jcarl43 has joined #gnucash
11:11:40 *** ChanServ sets mode: +v jcarl43
11:24:33 *** fell_laptop has quit IRC
11:24:33 *** fell has joined #gnucash
11:24:33 *** ChanServ sets mode: +o fell
11:39:51 <geneAum> chris-phone: yes, short selling. why would that be more clever? the borrow leg of a short sale IS a liability, no? at least, that's what's intuitive to me, but my accounting skills are pre-professional. :) The Balance Sheet report DOES seem to properly handle the "stock liability" as a "negative asset" in its determination of "Total Assets"
11:40:10 <geneAum> (why would tracking the liability as a negative asset be more clever?)
11:42:28 *** bertbob has quit IRC
11:48:18 *** jralls_laptop has joined #gnucash
11:50:24 <jralls_laptop> geneAum, The problem with your approach is that GnuCash determines whether an account is asset, liability, or equity by its type not its position in the chart of accounts. Type STOCK is a subtype of Asset so that's why your short sale shows up on the wrong side of the balance sheet.
11:51:02 *** bertbob has joined #gnucash
11:51:02 *** ChanServ sets mode: +v bertbob
11:51:11 <jralls_laptop> chris's suggestion to record a short sale as a negative balance is a work around for that limitation.
11:52:37 <jralls_laptop> CDB-Man_, an actual accountant, is best able to tell us what's correct in accounting for short sales, but intuitively I suspect that you're right and it's a design flaw in GnuCash.
11:52:49 *** lmat has quit IRC
11:53:17 *** Mechtilde has joined #gnucash
11:54:04 *** giuseppef has quit IRC
11:54:13 *** giuseppef has joined #gnucash
11:54:13 *** ChanServ sets mode: +v giuseppef
11:54:26 *** sbluhm has quit IRC
11:55:09 *** David has quit IRC
11:55:22 *** David has joined #gnucash
11:57:09 *** lmat has joined #gnucash
12:01:23 <jralls_laptop> chris, the multiple scheme targets in various CMakeLists.txt is how we get guile to build what we need in the right order. When guile encounters (use-modules foo) it first looks for foo.go on GUILE_COMPILED_LOAD_PATH, then for foo.scm on GUILE_LOAD_PATH and if it finds the latter it proceeds to compile it.
12:04:55 <jralls_laptop> It's pretty cumbersome to put all of the source directories containing scheme files in GUILE_LOAD_PATH so a long time ago someone (maybe gjanssens) set up the linking scheme which creates symlinks (or copies on Windows, which doesn't support symlinks) in the build dir.
12:06:23 <jralls_laptop> That was in autotools and codesmythe copied the paradigm to cmake when he wrote the conversion a few years ago. We really haven't come up with a better way so it's still there.
12:08:33 <jralls_laptop> Depending on how modules/submodules handle dependencies and auto-compilation we may be able to unify the targets--or may need to find a new workaround.
12:10:17 <jralls_laptop> chris, Is https://www.gnu.org/software/guile/manual/html_node/Modules.html the right documentation for what you're doing?
12:12:04 *** sbluhm has joined #gnucash
12:12:04 *** ChanServ sets mode: +v sbluhm
12:14:46 *** jervin has joined #gnucash
12:16:17 *** jervin has quit IRC
12:17:24 *** jervin has joined #gnucash
12:18:48 *** sbluhm has quit IRC
12:21:16 *** PowaBanga has quit IRC
12:22:15 *** Mechtilde has quit IRC
12:23:27 <CDB-PHONE_> geneAum jralls jralls_laptop: you are correct that the ideal proper treatment of short sale is a liability. that being said, tracking it in the software as a negative asset is what's known as a practical expedient. I can't imagine it being easy to make stock accounts flip to lability whenever they go negative. from a financial reporting perspective, reporting shorts net of long in the asset side is acceptable, provided the detailed breakdown is properly
12:24:02 <CDB-PHONE_> the short version is: while not 100% perfect, I don't have too many urgent bones to pick with presenting short sales as a negative asset
12:24:28 *** fell has quit IRC
12:24:38 <jralls_laptop> CDB-PHONE_ Thnaks.
12:24:45 *** fell has joined #gnucash
12:24:45 *** ChanServ sets mode: +o fell
12:24:59 <CDB-PHONE_> and unless a material portion of your portfolio is consisted of shorts, presenting it as net in assets would still be considered fair presentation
12:25:27 <CDB-PHONE_> what is material is a matter of professional judgement, but I can offer a "starting point" of let's say 25% of the portfolio
12:27:33 *** Mechtilde has joined #gnucash
12:29:55 *** codesmyt_ has joined #gnucash
12:30:55 *** o01eg has quit IRC
12:30:58 *** o01eg_ has joined #gnucash
12:31:08 *** codesmythe has quit IRC
12:32:50 *** o01eg__ has joined #gnucash
12:34:01 *** o01eg_ has quit IRC
12:37:23 <geneAum> CBD-Phone_ thanks indeed, very much appreciated
12:37:32 <CDB-PHONE_> as an anecdote, there are plenty of hedge funds based in the Cayman Islands that report under US GAAP, with sizeable short positions, and they report or net under the asset tally
12:37:51 <CDB-PHONE_> report it* under
12:37:57 <geneAum> interesting
12:38:07 <CDB-PHONE_> report it net under*
12:38:28 *** PowaBanga has joined #gnucash
12:47:22 *** guak has joined #gnucash
12:48:16 *** Agfarmer18 has joined #gnucash
12:50:19 *** CDB-Work has joined #gnucash
12:50:19 *** ChanServ sets mode: +v CDB-Work
13:06:31 *** sbluhm has joined #gnucash
13:06:31 *** ChanServ sets mode: +v sbluhm
13:09:19 *** giuseppef has quit IRC
13:09:28 *** giuseppef has joined #gnucash
13:09:28 *** ChanServ sets mode: +v giuseppef
13:11:46 *** sbluhm has quit IRC
13:13:57 *** giuseppef has quit IRC
13:14:08 *** giuseppef has joined #gnucash
13:14:08 *** ChanServ sets mode: +v giuseppef
13:15:33 *** jralls_laptop has quit IRC
13:16:05 *** jralls_laptop has joined #gnucash
13:17:04 *** jralls_laptop has quit IRC
13:17:07 *** jralls_laptop has joined #gnucash
13:18:01 *** jralls_laptop has quit IRC
13:21:39 *** jralls_laptop has joined #gnucash
13:21:55 *** sbluhm has joined #gnucash
13:21:55 *** ChanServ sets mode: +v sbluhm
13:24:39 *** giuseppef has quit IRC
13:25:08 *** giuseppef has joined #gnucash
13:25:08 *** ChanServ sets mode: +v giuseppef
13:40:10 *** Agfarmer18 has quit IRC
13:55:31 *** giuseppef has quit IRC
13:55:45 *** giuseppef has joined #gnucash
13:55:45 *** ChanServ sets mode: +v giuseppef
13:57:15 *** giuseppef has quit IRC
13:57:54 *** giuseppef has joined #gnucash
13:57:54 *** ChanServ sets mode: +v giuseppef
14:02:24 *** giuseppef has quit IRC
14:02:34 *** giuseppef has joined #gnucash
14:02:34 *** ChanServ sets mode: +v giuseppef
14:24:47 *** Agfarmer18 has joined #gnucash
14:26:19 *** Agfarmer19 has joined #gnucash
14:27:48 *** Agfarmer18 has quit IRC
14:43:00 *** PowaBanga has quit IRC
14:48:02 *** Agfarmer18 has joined #gnucash
14:50:53 *** Agfarmer18 has quit IRC
15:07:26 <gjanssens> jralls, chris: I think cstim originally came up with the linking scheme.
15:08:14 <jralls> OK.
15:08:28 <gjanssens> As for it being cumbersome to pull in all directories in GUILE_LOAD_PATH, in the current setup it certainly is.
15:08:53 <gjanssens> However if we'd restructure the sources a bit it could be doable.
15:09:11 <jralls> More like "a lot".
15:09:20 <gjanssens> Yeah, probably.
15:10:09 <gjanssens> It would help if the source structure of our guile sources already follows the guile module definitions in our files.
15:10:26 <jralls> Or vice versa. ;-)
15:10:30 <gjanssens> But we still have guile scattered over too many directories
15:11:11 <gjanssens> Vice versa is fine as well, but I think we need at least a gnucash base directory to separate our modules from guile's own or system installed modules.
15:11:25 <jralls> Of course.
15:12:09 <gjanssens> When I was thinking of this I was considering a gnucash directory in guile bindings and one in <repo-base>/gnucash
15:12:33 <gjanssens> And structure all our guile modules in those two like we'd want to install them.
15:12:56 <gjanssens> While building the guile load path should include these two directories
15:13:07 <warlord> Historically, the module names did follow the directory structure. We would make symlinks at the top level for 'gnucash', and then one could use-modules gnucash app-utils and it would work.
15:13:08 <gjanssens> At install time they would be merged into one.
15:14:15 <gjanssens> So only one directory needs to be added at install time, or actually even none as we install in guile's default module path for some time now already.
15:14:52 <gjanssens> Anyway just a fun thought experiment
15:15:35 <jralls> Well, one must still create a root `gnucash` dir to separate our modules.
15:15:50 <gjanssens> Of course
15:15:55 <jralls> We do that already, perhaps that's what you meant.
15:16:03 <gjanssens> Indeed
15:17:12 <jralls> Sequestering all of the guile code is kind of attractive, but I'm not sure that it plays well with the current way we do swig and have code that bounces back and forth between C and Scheme.
15:17:53 <gjanssens> An interesting question
15:18:46 <jralls> e.g. you were able to extract the core-utils and engine bindings to bindings/guile, but app-utils is still the blob it's always been.
15:19:01 <jralls> Was that because it was too hard or because you ran out of time?
15:19:50 <gjanssens> I held back for your c++-options work initially to not interfere too much with that and then I ran out of time.
15:20:08 <gjanssens> There are a few tricky bits but most should be solvable.
15:20:48 <jralls> Most should be solvable by rewriting them in C++. ;-)
15:20:50 <gjanssens> On tricky part is the expression parser. Most of it is written in C, but there's a section that allows running arbitrary scm commands.
15:21:07 <jralls> AKA a gaping security hole.
15:21:17 <gjanssens> Absolutely.
15:21:49 <jralls> The interest functions are written in Scheme, but they could be converted pretty easily.
15:21:58 <gjanssens> Unfortunately it's used. There was a discussion on the lists somewhere not too long ago on how to use custom interest payment functions.
15:22:59 <gjanssens> Well, about those interest functions, I suspect they actually come from C. There's a fin.c file with the exact same functions coming from some other gnome project
15:23:11 <gjanssens> What's their ancient excel variant called again ?
15:23:30 <jralls> I think that had to do with Jean Laroche's current balance variable PR that's been sitting since April.
15:24:20 <gjanssens> I guess so.
15:24:44 <gjanssens> Fin.c has a few standard functions, which are also available in the scm equivalent.
15:25:01 <gjanssens> However the scm equivalent later got extended with variations of these.
15:25:35 <jralls> Right. A drive-by PR about 5 years ago.
15:25:43 <gjanssens> True
15:27:36 <gjanssens> If the can-run-guile flexibility is to be retained in a C++ version of the expression parser, I would like to see a modular expression parser. That is, the base one handles things like +,-,*,... And the guile bindings provide an extension to the parser for guile function parsing.
15:28:42 <gjanssens> I didn't fully figure out how to do that while I was at it. Well, I didn't get anywhere really due to my lack of time.
15:28:56 <gjanssens> So I have ideas, but leave it there for now.
15:29:30 <gjanssens> s/there/at that/ ?
15:30:22 <gjanssens> warlord: FTR I haven't heard back from linas yet. I hope he got my mail and that google didn't dump it in his spam folder.
15:37:42 <jralls> ISTM we're chasing a tiny corner of the users here. We'd better serve more users by making the price-cell support a subset of standard spreadsheet functions than allowing a very few users to run arbitrary scheme code.
15:38:22 <jralls> And it's very few because that's how many scheme programmers exist in the world. Very few.
15:38:29 <warlord> gjanssens, I have not, either. He DID say he was busy.
15:42:24 <CDB-Work> gjanssens: unrelated, but i dont think we've formally "met" or "introduced" ourselves before, other than a couple of replies on the bugzilla
15:43:12 <CDB-Work> greetings, Geert -- I'm the CPA that's been having a bunch of very interesting conversations with jralls and chris about various accounting topics, as you probably saw with that giant ticket on the average cost not working
15:46:16 *** prg has quit IRC
15:46:36 <jralls> warlord, the post-recieve hook runs a program called $HOME/bin/update-main-website so code presumably has the ssh keys for the www.gnucash.org. Have you tried connecting from a code terminal session?
15:47:21 <jralls> If you can't get a terminal session maybe you could delete the master branch on www and re-push a clean one from code.
15:47:32 *** prg has joined #gnucash
15:58:58 <warlord> it doesn't ssh in. It does a port-knock and the server has a process listening that does a git pull.
16:11:46 *** Mechtilde has quit IRC
16:14:24 *** sbluhm has quit IRC
16:19:24 <gjanssens> jralls: thanks for being pragmatic. I agree dropping the ability for running arbitrary scheme code would be a reasonable way forward.
16:19:45 <gjanssens> CDB-Work: nice to meet you.
16:19:58 <gjanssens> I'm the core dev that ran out of time :(
16:20:51 <gjanssens> But it's certainly good for the project to have a CPA in the community to talk with on accounting matters
16:21:10 <jralls> gjanssens It'll be an easier sell if we say "replace" instead of "drop" ;-)
16:21:22 <gjanssens> :D
16:21:32 <gjanssens> replace with what ?
16:22:05 <jralls> spreadsheet-style formulas. Most everyone who'd use GnuCash knows how to use them.
16:22:41 <gjanssens> As for the gnucash-htdocs issue, the problem is the server side does a "git pull" rather than a "git fetch; git reset --hard origin/<branch>"
16:23:11 <gjanssens> If someone directly alters one of the files on the server, git pull sooner or later will fail due to a merge conflict.
16:23:18 <gjanssens> Apparently that is what happend now.
16:23:58 <gjanssens> Oh, now I see what you meant with supporting standard spreadsheet functions.
16:24:00 <gjanssens> Indeed
16:24:03 *** Mechtilde has joined #gnucash
16:24:06 <gjanssens> That would be nice
16:24:35 *** giuseppef has quit IRC
16:24:35 <gjanssens> The server update script is from our early git days still. We didn't have much experience yet.
16:24:50 <gjanssens> We don't use "git pull" anymore in our other automated scripts.
16:25:10 <jralls> Don't our other scripts push?
16:25:14 <gjanssens> But again it would be up to linas to fix that. The script is unreachable for me.
16:25:49 <gjanssens> Both the windows nightly build and the flatpak nightly build do automated updates of their clones
16:26:11 <jralls> But it seems really unlikely that anyone touched the www side directly. Only linas can do that and he wouldn't.
16:26:14 <gjanssens> I believe we use the fetch/reset scheme in both situations. If not, we definitely should
16:26:49 <gjanssens> I thought fell suggested that is exactly what linas has done somewhere in the past.
16:26:56 <gjanssens> But I may have misunderstood.
16:27:02 <jralls> The windows nightly pushes.
16:29:02 <gjanssens> Or it may have been a force-push by fell rather than linas.
16:30:32 <jralls> But nobody can push to www, nobody has the keys except linas, right?
16:31:36 <gjanssens> Yes
16:31:49 <gjanssens> I don't understand why you say windows nightly pushes ?
16:32:04 <gjanssens> The build script does a git pull on the gnucash-on-windows repository
16:32:28 <gjanssens> And jhbuild does a fetch/stash/reset/restore cycle
16:32:33 <jralls> Ah, I misunderstood which end of the script you meant.
16:34:07 <jralls> One can push to code.gnucash.org/gnucash-on-windows from the VM if one does an edit there, though I prefer to do changes on my local machine and push from there.
16:36:15 <jralls> Not that that has anything to do with the build scripts.
16:36:24 *** Aussie_matt has joined #gnucash
16:39:23 <gjanssens> So on the VM ssh is properly configured with an ssh key ? I didn't know that.
16:40:26 *** User_ has joined #gnucash
16:40:36 <gjanssens> That is, I assume to push from the VM it needs ssh to be set up fro git to be able to make the remote connection.
16:41:05 *** prg has quit IRC
16:44:24 *** prg has joined #gnucash
16:44:36 <jralls> It is, though not primarily for git: It needs an ssh key to scp the installer and log files back to code.
16:44:59 <gjanssens> Oh, right. Git can piggy-back on these settings.
16:46:29 *** User_ has quit IRC
16:46:35 <jralls> I just tried a push to gnucash-htdocs and got "NCAT: Connection reset by peer", nothing about a dirty working tree. Which makes sense, there's no git command in the hook.
16:48:21 <gjanssens> Yes, it's a port-knock. The actual command runs on the server. It doesn't seem to log back via ncat. I don't even know if ncat is capable of receiving output.
16:50:09 <jralls> It's not. I get the same results from www/beta, but the beta page updated and the main one didn't.
16:50:46 <jralls> So I wonder how fell concluded it's a dirty working tree.
16:58:38 <gjanssens> Time to go. Goodnight all!
16:59:30 <jralls> Goodnight!
16:59:30 *** gjanssens has quit IRC
17:02:40 <fell> jralls, I echoed the message, which warlord got while running www's pull.
17:03:04 *** giuseppef has joined #gnucash
17:03:04 *** ChanServ sets mode: +v giuseppef
17:07:50 <fell> There were uncommitted changes in news/news-script.php.
17:08:46 <fell> After I had added a line there in october, updates of www/master had stoped.
17:09:17 <fell> I see 2 possible causes:
17:12:55 <fell> 1. the previous change of that file was the application of Linas' php5 update. Probably Linas had applied it, too or had done some other change there. And then it was hidden until I tried to do the next change of the file.
17:13:22 <jralls> That update was 5 years ago.
17:14:07 <fell> 2. You had reset hard beta and AFAIK nobody has done it on www.
17:14:57 <fell> But nobody had tried to change the file in between.
17:28:27 <jralls> Hmm. If the pull script does a plain pull that would indeed do it. Oddly declaring $top_dir global doesn't seem to make any difference in the output so I can't tell if your news_script change got applied to master on www: There's no difference in the img urls between master and beta.
17:34:56 <warlord> jralls, git probably isn't copying stderr properly from the port-knock.
17:38:44 <jralls> Maybe, I'll look.
17:47:08 <jralls> git docs say that both stdout and stderr are forwarded. Is the ncat message on stdout?
18:03:59 *** codesmythe has joined #gnucash
18:04:00 *** ChanServ sets mode: +v codesmythe
18:04:39 *** codesmyt_ has quit IRC
18:04:45 <jralls> chris, Your maint-modularization is still failing CI. Could you change .github/workflows/ci-tests.yml line 53 from guile-2.2-dev to guile-3.0-dev or rebase on maint?
18:05:01 <chris> ok
18:06:45 <chris> in my branch, the deprecated scheme code was annoying me so they're gone
18:07:23 *** hussam has quit IRC
18:07:53 <jralls> I don't think that's what's failing. Have you looked at the CI dumps?
18:09:17 <chris> it's still not ready, lots of qif code to fix to be exoirtable
18:09:26 * chris busy left aem :)
18:50:35 <chris> the deprecated code was removed because I didnt know how to fix their modules
19:10:22 <jralls> Looks to me like you removed *deprecation* code rather than *deprecated* code... and you didn't really remove it, you just turned off testing it by removing the test target from CMakeLists.txt.
19:19:50 <chris> yes dont know how to fix it
19:22:21 *** prg has quit IRC
19:25:20 *** prg has joined #gnucash
19:27:35 <fell> I still get
19:27:37 <fell> From code.gnucash.org:gnucash-htdocs
19:27:38 <fell> Ncat: Connection reset by peer.
19:28:43 <jralls> fell, I think that's the normal output. I got that even when I updated beta and the update went through.
19:29:51 <fell> No, On the beta I got the full list of changed files with +++ at the end.
19:30:16 *** jcarl43 has quit IRC
19:33:07 <jralls> Oh, interesting.
19:33:18 <jralls> No ncat message at all?
19:45:26 <fell> jralls, see https://lists.gnucash.org/logs/2020/11/28.html#T23:16:49 for a successful commit.
19:47:13 <jralls> Ends up with the connection reset message. Hmm.
19:47:30 <fell> but full output.
19:49:08 <fell> On success I get the output, on error not.
19:50:00 <jralls> right. I wonder if git sends the stat output (that's what the list of files with +es and -es for adds and deletes is called) to stdout and the other stuff to stderr. In that case maybe adding 2>&1 to the knocker-script calls in gitolite/local_code/hooks/post_recieve would get it sent on.
20:15:41 <chris> I think qif-objects.scm needs to convert records to srfi-9 ... I'd delayed doing it because it's tedious
20:15:42 *** guak has quit IRC
20:25:54 *** storyjesse has joined #gnucash
20:25:54 *** giuseppef has quit IRC
20:47:11 <jralls> fell, I just pushed a commit to gitolite that redirects stderr to stdout for the knocker scripts. Maybe that will get the errors relayed. Unfortunately I agree with warlord that at this point only linas can fix it.
20:50:01 <warlord> I am HOPING he will respond (again) soon.
21:09:21 <CDB-Work> geneAum, jralls: further to our earlier conversation, i checked one of the hedge funds I work with that has a specific short selling strategy. in their case, the short sold shares was 2x the value of the long position shares, so in their case, given how large the short balance was, the shorts were presented on a gross basis in the liabilities secti
21:09:21 <CDB-Work> on
21:15:37 <fell> jralls, no success:
21:15:39 <fell> Repository git@code.gnucash.org:gnucash-htdocs
21:15:40 <fell> *** Mirror changes to origin (usually github)...
21:15:42 <fell> To ssh://github.com/Gnucash/gnucash-htdocs.git
21:15:43 <fell> d376b64..a2e08c3 master -> master
21:15:45 <fell> 3d08a84..d376b64 origin/master -> origin/master
21:15:46 <fell> *** Update main website...
21:15:48 <fell> From code.gnucash.org:gnucash-htdocs
21:15:49 <fell> Ncat: Connection reset by peer.
21:21:38 <CDB-Work> hmm, it is possible to make a flag in the balance sheet report, that when enabled, will SUM all of the credit balance STOCK accounts on the liabilities side
21:22:22 <CDB-Work> personally, I build my "formal" GAAP financial statements in Excel, and I would make the adjustment there. it's what we call a "topside entry" in accounting lingo
21:22:38 <CDB-Work> because it's done at the top and is temporal, rather than in the ledger and therefore permanent
21:23:20 <CDB-Work> that said, the nominal value of my short positions (including the fair value of options that I wrote, i.e. sold to open), is such a small percentage relative to the overall portfolio, that i dont bother
21:23:33 <CDB-Work> the notional exposure is a bit bigger, but even then, i dont bother
21:23:35 <warlord> fell, I don't know what to tell you. It's probably an interaction between ssh, git, and the script.
21:24:06 <CDB-Work> in the case of hedge funds, the fair value of the short securities is one factor -- another one is the notional value of the notional exposure on any derivatives such as options
22:01:10 <CDB-Work> https://www.iasplus.com/en/meeting-notes/ifrs-ic/not-added/2007/ias-39-short-trading -- hmm, interesting, when the question was posed to the IFRA committee a decade ago, a short sale was ruled as a liability, and its confirmed in paragraph 9 of IAS39 https://www.iasplus.com/en/standards/ias/ias39
22:01:45 <CDB-Work> though on a very practical level, for personal finances unless you are a major short trader, the practical expedient of just netting it is more than sufficient for personal finance
22:19:07 *** David has quit IRC
22:43:20 *** Robert8471 has quit IRC
22:43:50 *** Robert847 has joined #gnucash
23:03:17 *** Robert847 has quit IRC
23:27:01 *** PowaBanga has joined #gnucash