2018-09-16 GnuCash IRC logs
00:47:46 *** Mechtilde has joined #gnucash
00:59:32 *** greenshoe has joined #gnucash
01:01:10 *** greenshoe has quit IRC
01:01:15 *** greenshoe has joined #gnucash
01:10:26 *** greenshoe has quit IRC
01:39:17 *** shoonya has joined #gnucash
01:41:45 *** fell has quit IRC
01:42:35 *** fell has joined #gnucash
02:10:07 *** chf has quit IRC
02:10:41 *** chf has joined #gnucash
02:15:32 *** Mechtilde has quit IRC
02:22:16 *** chf1 has joined #gnucash
02:22:20 *** chf has quit IRC
03:10:47 *** greenshoe has joined #gnucash
03:13:47 *** greenshoe has quit IRC
03:47:49 *** storyjesse has joined #gnucash
03:52:46 *** storyjesse has quit IRC
03:58:21 *** boldstripe has joined #gnucash
04:28:46 *** ncv has joined #gnucash
04:36:03 *** fabior has joined #gnucash
04:37:08 *** boldstripe has quit IRC
04:39:07 *** fabior has quit IRC
04:51:31 *** fabior has joined #gnucash
04:58:59 *** boldstripe has joined #gnucash
05:05:21 *** fabior has quit IRC
05:08:32 *** bertbob has quit IRC
05:10:52 *** bertbob has joined #gnucash
05:13:58 *** bertbob has quit IRC
05:20:13 *** fabior has joined #gnucash
05:28:27 *** bertbob has joined #gnucash
05:31:17 *** boldstripe has quit IRC
05:35:10 *** chf1 has quit IRC
05:38:32 *** gncbot sets mode: +o fell
05:41:24 *** chf has joined #gnucash
05:42:45 *** shoonya has quit IRC
06:03:04 <chris> just had confirmation from #guile that it's worthwhile banishing (set!)... chapter 9.3.4 of guile manual states this is less optimizable
06:11:33 *** JayC has joined #gnucash
06:13:59 *** hmmwhatsthisdo has quit IRC
06:58:02 *** fabior has quit IRC
07:20:01 *** ncv has quit IRC
07:26:38 *** ncv has joined #gnucash
07:55:57 *** FoxT has quit IRC
08:02:03 *** FoxT has joined #gnucash
09:03:12 *** oozer has joined #gnucash
09:45:30 *** storyjesse has joined #gnucash
10:04:14 *** oozer has quit IRC
10:23:35 *** storyjesse has quit IRC
10:56:14 *** P-G has joined #gnucash
10:57:26 <P-G> Hi, is there a way to add a column to track transaction time?
11:00:47 <P-G> Also, I just set up a brokerage account following the website's guide but the bar at the bottom of the window shows the balance (of stock) as a unit of local currency, which it is not. Is there a way to change the unit or value to properly correspond?
11:06:06 *** oozer has joined #gnucash
11:17:32 *** O01eg has quit IRC
11:21:57 *** O01eg has joined #gnucash
11:28:02 *** O01eg has quit IRC
11:55:36 *** O01eg has joined #gnucash
11:55:44 *** boldstripe has joined #gnucash
12:08:06 *** O01eg has quit IRC
12:14:53 *** boldstripe_ has joined #gnucash
12:15:10 *** boldstripe has quit IRC
12:15:10 *** boldstripe_ is now known as boldstripe
12:21:49 *** O01eg has joined #gnucash
12:27:31 <jralls> P-G: No. Time of day is not significant in accounting.
12:29:34 *** gjanssens has joined #gnucash
12:29:35 *** ChanServ sets mode: +o gjanssens
12:29:42 <gjanssens> .
12:30:08 <jralls> P-G: What version of GnuCash? My 3.2 has the status bar in stock accounts showing shares and current value (based on the last-retrieved price).
12:32:41 <jralls> gjanssens: I've managed to wedge my Windows build so that guile compilation is wedged. load-extension errors out with "Value out of range 0 to 4294967295: 1744408812"
12:32:56 *** boldstripe has quit IRC
12:33:48 <jralls> gjanssens: The only thing that makes sense as cause is running `pacman -Syuu` as part of build_package.ps1 yesterday.
12:36:31 <gjanssens> jralls: that's bad news
12:37:30 <jralls> Yeah. Derek's VM isn't yet affected... Hmm, I'll look at its pacman.log.
12:37:32 <gjanssens> I still haven't recovered my Windows box since I had to repurpose it as asterisk phone central.
12:37:48 <gjanssens> So I can't experiment with you here :(
12:38:32 <jralls> Why not set up a Windows VM on your development machine?
12:42:08 <jralls> Hmm. Daily updates to MSYS2 on the VM, so it shouldn't be any different from mine.
13:09:46 <gjanssens> Why not set up a VM? Well, always more interesting things to do first...
13:10:23 <jralls> Perhaps the reason the windows box is still handling your phones as well? ;-)
13:20:45 *** Mechtilde has joined #gnucash
13:24:42 *** Mechtilde has quit IRC
13:25:20 *** Mechtilde has joined #gnucash
13:29:07 <gjanssens> jralls: indeed :D
13:31:13 *** bertbob has quit IRC
13:31:47 *** Mechtilde has quit IRC
13:33:45 *** bertbob has joined #gnucash
13:36:06 <P-G> Jralls: "GnuCash 2.7.4 development version Build ID: git 2.7.4+ (2018-02-03)". It's the most recent stable from my repository.
13:37:12 *** fell has quit IRC
13:37:33 *** fell has joined #gnucash
13:37:39 <jralls> P-G: Change distros. Not only is that a year old, it's an unstable development build that shouldn't even be in a distro's package manager.
13:38:12 <jralls> P-G: Fedora seems to be doing the best job of staying up to date right now.
13:42:47 <P-G> Hm. If I install from source, would I have to check for updates manually?
13:43:39 <P-G> Or maybe I would be better off using an older stable version from my repo (if I can find one)?
13:49:22 <jralls> P-G: You'd be better off using a distro that if it ships unstable betas it keeps up with updating them regularly. If they've done that with GnuCash who knows what other unstable stuff is on your system?
13:50:22 <jralls> P-G: But to answer your question directly, yes, you'd have to check for updates manually. But that would still be better than what you have.
13:53:29 <P-G> Ok, I'm going to try asking around about this before I consider switching distros. It is definitely worrisome, but there may be something I've overlooked. It seems like there were older (stable) versions available at one point, but they aren't listed anymore. Maybe they're still available.
13:55:08 <P-G> About the time column, I suppose that makes sense. I feel like it's worth including as an option though, or at least a generic text string column that users can name freely.
13:56:12 <jralls> P-G: You're missing the point. What if they've also installed an obsolete unstable version of WebKit with a significant CVE? That distro isn't trustworthy.
13:57:53 <P-G> No, I get that. The problem is that I do trust them, so I feel like I just did something wrong. If there isn't a good explanation for this, I will definitely consider switching distros.
14:02:49 <P-G> Hypothetically, if GnuCash is using a relational database, could I add a time column manually to the database table and write a trigger to operate on that column or would this be a bad idea?
14:03:28 <P-G> I do realise that probably wouldn't appear in GnuCash.
14:04:19 *** boldstripe has joined #gnucash
14:05:37 <jralls> Hypotheically, if GnuCash's SQL backend was used as more than just an object store you might. As it is, GnuCash loads all of the data at the beginning of the session and only writes to it when something changes, but never reads it again during the session.
14:07:07 <jralls> If you want to add information for processing with some other tool e.g. piecash then it would be better to add a separate table with the transaction guid and whatever extra fields you want.
14:09:27 <jralls> Note as well that GnuCash *does* use timestamps (datetime in MySql because MySql timestamps can't do time after January 2038) for posted date, but they're set to 10:59 UTC to maintain their date identity across most timezones.
14:12:14 <P-G> Ok, that's good to know. I'll try adding a table if I do end up feeling the need for transaction times.
14:13:07 <P-G> Just out of curiosity, is the plan to eventually use SQL in a real-time way?
14:17:53 <jralls> Eventually, but we have to change the engine so that it doesn't depend on having everything in memory as it does now. That leads us to better construction and destruction, which puts the C++ conversion in front of changing to always having a SQL backend and using XML as a backup format only.
14:18:05 <jralls> So it's several years out.
14:19:48 <P-G> Good to know. Maybe I will offer to help if I manage to learn C++ before then. :p
14:22:50 *** frakturfreak has joined #gnucash
14:38:03 *** gncbot sets mode: +o fell
14:47:38 *** oozer has quit IRC
14:47:54 *** KevinDB has quit IRC
14:52:06 <P-G> So apparently they marked 2.7.4-r1 stable because 2.6 was still using the "years-abandoned security-hole-ridden webkit-gtk:2" and 3.2 was not out yet. As for why they haven't yet adopted 3.2, there just hasn't been any discussion about it until now and they assume a certain bug from 3.1 still exists in 3.2 (not sure why).
14:53:56 *** KevinDB has joined #gnucash
14:55:01 <P-G> Do you know about this "unnecessary parentheses in declaration of __gtk_reserved1"? https://bugs.gentoo.org/655934
15:03:45 *** boldstripe has quit IRC
15:10:52 *** boldstripe has joined #gnucash
15:11:27 *** ncv has quit IRC
15:18:46 <fell> chf, I think the wiki is the better place for your script than bugzilla.
15:21:15 <chf> I didn't know where to put it, but wanted to document it, and of course I hope that somebody with XSLT knowledge can improve the original pieco of software.
15:21:30 <chf> o -> e
15:25:57 <chf> For each additional element to change from defaults or filter out of the template, we'll get more "sed" or "grep" statements; and the current version of "gnc2xea.xsl" cannot be distributed in Gnucash's "contrib" directory (at least not without such a script, which must then be generalised).
15:26:00 *** boldstripe has quit IRC
15:36:48 <fell> Perhaps we should start a new page about the xslt, linked to Account Hierarchy Template
15:38:54 <fell> I was considering the color not set slots from your templates and eventually other slots with default values.
15:39:19 <fell> ... to remove ...
15:43:58 <chf> I've just added the script to the wiki as well, but the page is a bit confusing in its current state, I think.
15:44:52 *** boldstripe has joined #gnucash
15:45:26 *** greenshoe has joined #gnucash
15:45:27 <chf> Not mixing up the explanation for manual conversion with the script based approach is a good idea.
15:50:45 <fell> yes, it is becoming too big. I think the scripts incl. their description should get their own pages.
15:50:47 <chf> We should definitely make a general wrapper script which takes ONLY the 4 necessary parameters: 1) file name of the file to convert; 2-4) title, short, long descriptions.
15:50:54 *** boldstripe has quit IRC
15:51:51 <chf> And that should use as few external programmes as possible: I'd like to avoid "sed", "grep", "awk"…
15:52:22 <chf> This requires to put the additional changes inside the xslt part.
15:57:17 <chf> I've looked into "gnc2xea.xsl", but could not figure out how to make my desired changes. I DO know, however, how to replace fixed strings inside a line with 'sed -e "s/…/…/g"', and remove whole lines with "grep -v …" for example.
15:58:33 <fell> The search is done by <xsl:template match=
15:58:53 <fell> and inside is the replacement
16:02:34 <fell> https://de.wikipedia.org/wiki/XSL_Transformation is a good starting point.
16:03:14 <jralls> P-G: Apparently Gentoo has lost everyone with anything at all between their ears. As the poster notes, it's a fixed gtk bug, but apparently nobody cares to address that. To answer Aaron W. Swenson's idiotic question, no, of course it isn't fixed in GnuCash 3.2 because it's not in GnuCash code. But it raises an interesting question: How are they building Gtk3 at all?
16:09:09 *** frakturfreak has quit IRC
16:12:37 *** fabior has joined #gnucash
16:28:57 <P-G> Thanks for answering. I like Gentoo's ethos, documentation, and community. Maybe it's because I don't use the bug tracker or forums, but I haven't run into many people like Swenson. That said, this is pretty concerning. Specifically, that the guy I talked to doesn't seem to have read the bug he cited.
16:35:37 <P-G> Arch has some good documentation too, and I used that for a few years, but people would always tell me to switch software whenever I had a question... "This is how you do it with NetworkManager", "it just works with PulseAudio", etc. If I'm trying to understand how something works, that just doesn't help.
16:36:38 <P-G> I use Debian on servers, and that works well. I used to use it on my P.C. too, maybe I will again.
16:41:47 <fell> chf: the header handling of our xslt seems to miss a few <br /> e.g. in the namespace section.
16:42:26 <P-G> Installing 3.2 did fix that little value anomaly. :)
16:53:21 <jralls> P-G: Debian so far supports GnuCash3 only in unstable. Stable has 2.6.15 and testing has no GnuCash at all.
16:53:48 *** oozer has joined #gnucash
16:54:08 <jralls> P-G: Good that you finally got GnuCash 3.2 installed, though. Just in time for 3.3 to release in 2 weeks!
16:55:53 *** fabior has quit IRC
17:15:01 <fell> Good night!
17:48:37 *** fell has quit IRC
18:03:01 *** gjanssens has quit IRC
18:10:14 *** greenshoe has quit IRC
18:15:55 *** fell has joined #gnucash
18:18:59 *** greenshoe has joined #gnucash
18:52:26 *** chf has quit IRC
18:53:41 *** chf has joined #gnucash
18:54:45 *** pilotauto has joined #gnucash
19:15:02 <chris> #316 and #411 both modify net-charts.scm and will be in conflict. Any clue which one would be mergeable first?
19:16:11 <chris> P-G 3.2 unfortunately has 1 pesky html-utilities.scm bug which is fixable with a modicum of hacking https://bugs.gnucash.org/show_bug.cgi?id=796696 or wait for v3.3
19:19:10 *** greenshoe has quit IRC
19:21:10 *** greenshoe has joined #gnucash
19:37:09 <jralls> chris: If we merge 316 and you redo 411 on top of it will that take care of the ugly/slow barchart code?
19:38:44 <jralls> chris: And is the barchart code in 411 ugly, slow, or both?
19:49:53 <P-G> Lol, that's my luck. I'm glad the development is going well though. Thanks for the link, Chris; I probably will be on 3.2 for quite a while.
19:50:00 <chris> 316 is chartjs upgrade, 411 is optimize-balance so they'll both work, i'll just need to know which one will be first
19:50:48 <P-G> I'm signing off for now, but I'm sure I'll be back sooner or later. Thanks for the advice. :)
19:51:09 *** P-G has quit IRC
19:51:10 <chris> it's category-balance.scm which can't be reoptimized yet
19:52:27 <chris> i'll need a few hours to clean up 316 completely, separating chartjs-upgrade from code-cleanup
19:55:45 <chris> category-balance and net-charts were 'optimized' by p.broadbery 3? years ago in a difficult way, by caching certain balance calculations... net-charts was easy to deoptimize/reoptimize because xaccAccountGetBalanceAsOfDate is a direct replacement for the slow qof-query balance calculator
19:56:24 <chris> but category-balance can't be deoptimized/reoptimized yet so I've maintained p.broadbery's strategy for now
19:57:20 *** greenshoe has quit IRC
19:58:29 *** jethrob has quit IRC
19:59:52 *** jethrob has joined #gnucash
20:02:51 *** JayC has quit IRC
20:05:01 *** fell has quit IRC
20:47:40 *** JayC has joined #gnucash
21:11:36 <chris> i'll need to separate cleanup vs chartjs work in 316
21:57:42 *** greenshoe has joined #gnucash
21:58:17 *** JayC has quit IRC
21:59:36 *** storyjesse has joined #gnucash
22:00:42 *** greenshoe has quit IRC
22:03:20 *** JayC has joined #gnucash
22:24:44 *** Agfarmer18 has joined #gnucash
22:29:20 *** oozer has quit IRC
22:47:20 <chris> jralls I suggest 411 is merged first... I think there's a lot of work in 316 cleanup that I'm not sure will be realistic for 3.3
22:52:34 <chris> the problem is jqplot and chartjs are not 1-to-1 compatible and there were a lot of small adjustments i've made along the way; i can deconstruct #316 but it will take time
23:01:21 <chris> https://groups.google.com/forum/#!topic/jqplot-dev/tDkY2z0q2t4
23:17:22 *** Agfarmer18 has quit IRC
23:25:28 *** greenshoe has joined #gnucash
23:29:01 *** greenshoe has quit IRC
23:29:32 *** greenshoe has joined #gnucash
23:32:32 *** greenshoe has quit IRC
23:35:50 *** greenshoe has joined #gnucash
23:45:22 *** CDB-Away_ has quit IRC
23:51:24 *** greenshoe has joined #gnucash