2017-04-17 GnuCash IRC logs
00:20:31 *** nitin-mobile has quit IRC
00:50:34 *** mlncn has joined #gnucash
01:03:52 *** gr8kodr has joined #gnucash
01:07:37 <gr8kodr> still trying to understand the Windows version ... why is it that I can't enter my dates as 'datetime' values, but if I look at the XML of the data file, the values I enter are stored in 0000-00-00 00:00:00 +0000 notation?
01:07:37 <gncbot> gr8kodr: Sent 21 hours and 23 minutes ago: <fell> Something like in http://code.gnucash.org/docs/C/?
01:10:12 <gr8kodr> fell: ty for that link, I didn't realise there were compiled versions available :) that was exactly what I was after with regards the Docs
01:24:02 <fell> gr8kodr: It is atimestamp in ISO format.
01:33:19 *** fell_ has joined #gnucash
01:35:11 *** fell has quit IRC
01:50:14 *** O01eg has quit IRC
01:52:21 <gr8kodr> fell: I want the timestamps in ISO format, but within the Windows Interface, there is no way to enter timestamps, only dates
01:54:02 <gr8kodr> I don't understand why if ISO format is selected as the date format in Options, the Interface doesn't allow ISO format data entry
01:55:01 <gr8kodr> is this just a Windows trait, or is it the same for Linux and Mac?
03:36:17 *** Mechtilde has joined #gnucash
04:07:47 *** Mechtilde has quit IRC
04:31:31 *** fell_ is now known as fell
04:31:40 *** gncbot sets mode: +o fell
04:34:04 <fell> gr8kodr: The interface only accepts dates, no time. That led to confusion already before.
04:35:48 <gr8kodr> fell: so only way to 'preserve' correct transaction time is to close GC and edit in raw XML?
04:38:16 <gr8kodr> .. or possibly, save as ... to MySQL, edit the datetimes there, then save as .. back to xml
04:42:54 <gr8kodr> at present, the dates I enter are being converted to local tz and having time stripped before being saved with a pre-defined time and tz offset ...
04:46:32 <gr8kodr> I would like to be able to save as UTC rather than locale ... my understanding is that if I export to MySQL, MySQL will save as UTC and display back according to the 'session'.'timezone' value
04:48:21 <gr8kodr> for the purpose of saving my pricedb, this is exactly what I need, as the prices I have are all quoted as UTC
04:49:57 *** fabior has joined #gnucash
04:51:39 <gr8kodr> once I have my priceDB populated, the no time issue becomes irrelevant to my purposes :)
05:12:32 *** Mechtilde has joined #gnucash
05:12:37 *** gjanssens has joined #gnucash
05:12:37 *** ChanServ sets mode: +o gjanssens
05:12:45 <gjanssens> .
05:12:45 <gncbot> gjanssens: Sent 13 hours and 30 minutes ago: <jralls> I finally got webkit to rebuild and I've reverted the kluge. Delete the webkit directory to install the new libs, and note that opensp also has a dependency on pthreads so one may need to force a rebuild on that too.
05:28:08 *** Mechtilde has quit IRC
05:35:51 <gjanssens> jralls: thanks for the work on webkit. I'll start a build locally as well to see how it goes.
05:37:46 <gjanssens> I noticed you didn't increase the build version (from -1 to -2). Perhaps we should to better distinguish the new tarball from the old one (of which several copies may still circulate outside of our control)
06:00:40 *** To7 has quit IRC
06:12:21 *** matthew has joined #gnucash
06:13:04 <matthew> hello all, I'm running mageia linux, the version in our repo's is 2.6.5 is that considered too old to use? or am I good to go?
06:14:43 *** fabior has quit IRC
07:39:59 *** User_ has joined #gnucash
07:42:39 *** User__ has joined #gnucash
07:44:40 *** User_ has quit IRC
07:50:05 <gjanssens> matthew: That version is almost 2.5 years old and several bugs have since been discovered and fixed.
07:50:33 <gjanssens> You can check the release notes on the gnucash website to check if there are any affecting how you intend to use gnucash
07:50:44 <gjanssens> For basic usage it's probably ok.
07:51:11 <gjanssens> However if you run into an issue, the first thing you'll probably be asked is to upgrade to a more recent version :)
07:51:28 * gjanssens has to reboot so I'll be offline for a while.
07:51:29 <gjanssens> BRB
07:51:51 *** gjanssens has quit IRC
07:52:45 *** rickoehn has joined #gnucash
07:53:40 *** gjanssens has joined #gnucash
07:53:40 *** ChanServ sets mode: +o gjanssens
08:11:57 <matthew> gjanssens: hi, thanks for response!
08:12:14 <matthew> Yeah, it is pretty old looking at the releases since :(
08:12:41 <matthew> I haven't looked to far into it, is there a docker image/vm or other?
08:49:00 *** storyjesse has joined #gnucash
09:09:13 *** warlord has joined #gnucash
09:09:14 *** gncbot sets mode: +o warlord
09:14:25 *** Mechtilde has joined #gnucash
09:22:33 <gjanssens> matthew: no idea for mageia
09:23:07 <gjanssens> Have you contacted their gnucash maintainer to learn why their package is so far behind ?
09:34:46 <warlord> .
09:51:57 *** kael has joined #gnucash
10:09:02 *** matthew has quit IRC
10:24:15 *** To7 has joined #gnucash
10:45:47 *** Jacques has joined #gnucash
10:47:44 *** KaiForce has joined #gnucash
10:49:39 *** Jacques has quit IRC
10:55:28 *** Mechtilde has quit IRC
11:18:58 *** chf1 has joined #gnucash
11:19:43 *** chf has quit IRC
12:02:07 *** kael has quit IRC
12:08:47 *** kael has joined #gnucash
12:17:02 *** storyjesse has quit IRC
13:00:59 *** kael has quit IRC
13:05:45 <jralls> gjanssens: I did change the tarball version, from webkit-1.8.3-minGW.tgz to webkit-1.8.3-minGW-1.tgz. I didn't bump the webkit version because I didn't touch the other tarball.
13:07:20 <jralls> gr8kodr: Time of day is not meaningful in accounting. The internal representation is a 64-bit int because that's how standard C deals with dates, but that's an implementation detail.
13:13:48 *** kael has joined #gnucash
13:17:41 <gjanssens> jralls: oh, now I see it! You did this as part of reverting the clutch. I missed that.
13:18:09 <gjanssens> By the way, I'm having some trouble getting master to build on WinXP
13:18:23 <jralls> Is "clutch" Dutch/Flemish for "kluge"?
13:18:43 *** kael has quit IRC
13:19:05 <jralls> What kind of problem on XP? Not that I'm overly surprised, the XP mswincrt probably doesn't support C++11.
13:20:23 <gjanssens> https://paste.fedoraproject.org/paste/-~wH-ktce7~QrjuhrdJE6F5M1UNdIGYhyRLivL9gydE=
13:20:37 <gjanssens> It seems to be not finding _stricmp
13:21:14 <gjanssens> which is used in googletest headers
13:21:34 <gjanssens> A quick search on the net suggests this should be available on WinXP though
13:21:37 <gjanssens> http://xpdll.nirsoft.net/msvcrt_dll.html
13:22:12 <gjanssens> I haven't had time to investigate very thoroughly as I'm preparing for a meetup with my accountant on thursday
13:23:26 <gjanssens> as for my use of "clutch" that was simply a non-native mistake.
13:24:04 <gjanssens> I meant kludge (which appears to be a synonym for kluge http://www.dictionary.com/browse/kluge)
13:25:54 <jralls> Ah, I had that problem on Win7 yesterday. It's a problem with Marshall's goofy partial-include mechanism. I got it to finish by commenting out the guard at line 69 of mingw/include/strings.h and the corresponding #endif. I still need to find a more elegant solution.
13:26:39 <jralls> BTW, there's another bug in the same mechanism in stdio.h at line 99: He used || and meant &&.
13:27:35 <jralls> "if !(foo || bar)" almost never does what the writer intends!
13:31:26 <warlord> jralls: it depends what the writer intended
13:31:53 <gjanssens> jralls: ok. It looks though I have a different version of these files because the line numbers don't match
13:32:29 <gjanssens> Oh, nevermind
13:32:50 <gjanssens> It's a visual offset in the line numbering of kwrite on WinXP
13:33:00 <gjanssens> The line numbers themselves are correct
13:33:10 <jralls> warlord: Of course. More correctly, the author seldom wants what !(foo || bar) means.
13:33:40 <jralls> gjanssens: GASP! I thought you were an emacsen!
13:33:52 <gjanssens> nope :)
13:34:43 <gjanssens> And still somthing is different.
13:34:52 <gjanssens> Line 99 is a closing comment on my system
13:35:25 <gjanssens> What's the content of the line having the issue in stdio.h ?
13:35:44 <jralls> Sorry, it's line 199.
13:36:06 <gjanssens> ah, ok
13:37:02 <jralls> It's #if ! (defined _STDIO_H || defined _WCHAR_H)
13:37:32 <jralls> and should be #if ! (defined _STDIO_H && defined _WCHAR_H).
13:37:35 <gjanssens> Interestingly it's #if ! (defined _STDIO_H && defined _WCHAR_H) here
13:37:45 <gjanssens> And I didn't change this myself
13:38:04 <warlord> jralls: could be. But it's certainly shorter than writing !foo && !bar ;)
13:38:29 <jralls> gjanssens: That *is* interesting.
13:39:04 <jralls> warlord: No, it's one character longer, two if you use the unnecessary space like Marshall did.
13:39:38 <warlord> I tend to use that construct when there are more than two items.
13:39:50 <gjanssens> jralls: do you know offhand what package provides this file ?
13:40:00 <gjanssens> to compare versions
13:41:24 <jralls> gjanssens: Should be part of the development core, but I don't know exactly how the packages are broken down and AFAIK mingw-get doesn't have a "provides" command.
13:43:15 <gjanssens> Indeed. I use grep on mingw/var/lib/mingw-get/data which has the installation catalogs
13:43:32 <jralls> No, wait. It's going to be MINGW_W32_API, which I upgraded from 3.17-2 to 3.18-2 to correct another problem with gtest.
13:43:41 <gjanssens> Appears stdio.h comes from mingwrt-3.22.4
13:44:24 <jralls> Which is the other package I upgraded in that same commit, 25e36b.
13:47:06 <gjanssens> According to the install manifest I have both packages installed on the most recent versions
13:47:22 <gjanssens> I'll remove them and reinstall to see what will happen
13:47:39 <jralls> BRB
13:49:21 <gjanssens> k
13:50:52 *** kael has joined #gnucash
13:53:42 <gjanssens> jralls: used mingw-get remove to remove the two packages so they got reinstalled (from download cache) on the next invocation of install.sh
13:53:58 <gjanssens> stdio.h is still having the correct invocation (using &&)
13:54:21 <gjanssens> strings.h still needs patching to include _stricmp
13:55:30 *** fekepp has joined #gnucash
13:56:54 <jralls> gjanssens: Hmm. So stdio.h got borked on my system somehow. Interesting.
13:58:12 <gjanssens> Removing the guard in strings.h does allow me to continue the build
13:58:26 <gjanssens> No need on my system to alter stdio.h
14:02:22 *** kael has quit IRC
14:05:37 *** frakturfreak has joined #gnucash
14:15:16 *** O01eg has joined #gnucash
14:15:18 *** kael has joined #gnucash
14:37:07 <jralls> gjanssens: I found the real problem with strings.h. It's called by string.h which delays setting __STRING_H_SOURCED__ to after the #if _EMULATE_GLIBC_ #include <strings.h>. That's enabled by using -std=gnu++11 instead of -std=c++11.
14:38:02 <jralls> Moving #define __STRING_H_SOURCED__ up a few lines also allows the build to complete.
14:42:34 <jralls> Ah, I see that I overrode that with a CFLAG in configure.ac but not in CMakeLists.txt (2b4137c).
14:51:40 <gjanssens> Good detective work!
14:52:13 <jralls> Fix pushed.
14:55:09 *** User_ has joined #gnucash
14:55:35 *** User__ has quit IRC
14:55:41 *** cyphase has quit IRC
14:55:48 *** mrklintscher has quit IRC
14:55:52 *** User_ has quit IRC
14:55:52 *** KaiForce has quit IRC
14:55:52 *** To7 has quit IRC
14:55:52 *** warlord has quit IRC
14:55:52 *** gr8kodr has quit IRC
14:55:52 *** ThomasKeller has quit IRC
14:55:52 *** jonas has quit IRC
14:55:52 *** miklcct has quit IRC
14:55:52 *** ajm has quit IRC
14:55:52 *** jchonig has quit IRC
14:55:52 *** mikee has quit IRC
14:55:52 *** CDB-Away_ has quit IRC
14:55:52 *** karelk has quit IRC
14:55:52 *** GabrieleV has quit IRC
14:55:52 *** puck has quit IRC
14:55:52 *** xmaka has quit IRC
14:56:01 *** cyphase has joined #gnucash
14:56:01 *** User_ has joined #gnucash
14:56:01 *** KaiForce has joined #gnucash
14:56:01 *** To7 has joined #gnucash
14:56:01 *** warlord has joined #gnucash
14:56:01 *** gr8kodr has joined #gnucash
14:56:01 *** ThomasKeller has joined #gnucash
14:56:01 *** jonas has joined #gnucash
14:56:01 *** miklcct has joined #gnucash
14:56:01 *** ajm has joined #gnucash
14:56:01 *** jchonig has joined #gnucash
14:56:01 *** mikee has joined #gnucash
14:56:01 *** CDB-Away_ has joined #gnucash
14:56:01 *** karelk has joined #gnucash
14:56:01 *** GabrieleV has joined #gnucash
14:56:01 *** puck has joined #gnucash
14:56:01 *** xmaka has joined #gnucash
14:56:01 *** irc.gimp.ca sets mode: +oo warlord mikee
14:56:15 *** mrklintscher has joined #gnucash
14:56:58 *** gjanssens has quit IRC
15:05:36 *** fabior has joined #gnucash
15:16:56 *** gjanssens has joined #gnucash
15:16:56 *** ChanServ sets mode: +o gjanssens
15:17:37 *** kael has quit IRC
15:27:37 *** chf1 has quit IRC
15:34:56 *** chf has joined #gnucash
15:36:06 <gjanssens> Back... my computer had locked up...
15:36:17 <gjanssens> jralls: saw your push. I'll test it later
15:36:31 <gjanssens> (though no doubt it works :)
15:37:03 *** kael has joined #gnucash
16:12:35 <jralls> gjanssens: Well, it worked for me...
16:15:45 *** O01eg has quit IRC
16:29:07 *** O01eg has joined #gnucash
16:34:26 *** frakturfreak has quit IRC
16:38:46 *** chf has quit IRC
16:38:55 *** chf has joined #gnucash
16:53:20 <jralls> gjanssens: So on to mingw64. I'm inclined to do the boot script in PowerShell instead of VBS, and I'm inclined not to do anything special to support anything earlier than Win7.
16:54:58 <jralls> gjanssens: But you're just now building on XP so I wonder if I should tip in a different direction.
16:59:40 <jralls> gjanssens: OTOH I'm just reminded of the security implications of using PowerShell; we'll need a signing certificate and I suspect that my Apple one won't do.
17:10:13 *** User_ has quit IRC
17:37:13 <gjanssens> jralls: I intended to do the mingw64 work on a Win7 virtual machine anyway, so I'm fine dropping WinXP as build system
17:38:09 <gjanssens> I'm not aware of the PowerShell security implications you refer to. Perhaps we can talk about this later. I'm off to bed now.
17:40:02 *** gjanssens has quit IRC
18:01:37 <jralls> @tell gjanssens The security issue is that PowerShell won't run scripts by default and only the least-restrictive Windows setting allows it to run unsigned scripts that didn't originate on the local computer.
18:01:37 <gncbot> jralls: The operation succeeded.
18:02:19 <jralls> @tell gjanssens Woo-hoo! Msys2 has jhbuild!
18:02:19 <gncbot> jralls: The operation succeeded.
18:11:58 *** meb has joined #gnucash
18:14:47 *** fabior has quit IRC
18:16:18 *** meb has quit IRC
18:17:03 *** meb has joined #gnucash
19:37:47 *** warlord has quit IRC
19:48:40 *** warlord has joined #gnucash
19:48:40 *** gncbot sets mode: +o warlord
20:04:06 *** kael has quit IRC
20:18:12 *** cyphase has quit IRC
20:33:08 *** warlord has quit IRC
20:36:48 *** warlord has joined #gnucash
20:36:48 *** gncbot sets mode: +o warlord
21:39:35 *** cyphase has joined #gnucash
21:54:27 *** warlord has quit IRC
22:06:15 *** cyberspectre has joined #gnucash
22:07:56 *** User_ has joined #gnucash
22:13:40 *** meb has quit IRC
22:14:35 *** meb has joined #gnucash
22:17:51 *** kael has joined #gnucash
22:27:07 *** kael has quit IRC
22:27:10 *** gncbot has joined #gnucash
22:28:56 <cyberspectre> Hi everyone. Does anyone successfully use gnucash direct-connected to a chase account?
22:39:38 *** User_ has quit IRC
23:26:36 <gr8kodr> jralls: there is one instance where time of day is meaningful, that seems to be overlooked in every software I've seen ... Running Balances - a transaction taking place at 5AM shouldn't appear before a transaction taking place at 7PM on the same day ...
23:28:19 <gr8kodr> transactions recently imported from CSV are not necessarily going to have 'numbers' in sync with the book, but the time the transaction occurred will still enable correct ordering on a given day
23:31:48 <gr8kodr> additionally, time does come into effect when dealing with FX quotes .... if my quotes are all for 1700 UTC, and I enter a transaction that occurred at 1300 UTC, then the program should use yesterdays provided quote, rather than the quote for today which isn't in effect yet