2014-06-12 GnuCash IRC logs
00:07:20 *** isaacd has quit IRC
00:17:22 *** StuM has quit IRC
00:18:26 *** StuM has joined #gnucash
00:20:15 *** fell__ is now known as fell
00:21:45 <fell> retrospectacus: xml is the well tested traditional format.
00:35:20 *** MechtiIde has quit IRC
00:44:35 *** gjanssens has joined #gnucash
00:44:35 *** gncbot sets mode: +o gjanssens
00:59:55 *** floh1111 has joined #gnucash
01:07:16 *** Krzysiek_K has left #gnucash
01:53:09 *** StuM1 has joined #gnucash
02:00:35 *** StuM has quit IRC
02:51:44 *** fell has quit IRC
03:13:18 *** GabrieleV has quit IRC
03:13:51 *** GabrieleV has joined #gnucash
03:29:45 *** Bodhi-Baum has joined #gnucash
04:06:10 *** fell has joined #gnucash
04:06:11 *** gncbot sets mode: +o fell
04:06:32 *** floh1111 has quit IRC
04:23:09 *** fell_ has joined #gnucash
04:23:09 *** gncbot sets mode: +o fell_
04:28:20 *** fell has quit IRC
04:31:30 *** aqua___ has joined #gnucash
05:09:12 *** Bodhi-Baum has quit IRC
05:17:39 *** himaxx has joined #gnucash
05:20:16 *** himaxx has quit IRC
05:23:03 *** Jimraehl1 has joined #gnucash
06:26:02 *** jmd has joined #gnucash
06:32:58 *** jmd has quit IRC
06:32:58 *** jmd has joined #gnucash
06:50:37 *** jmd has quit IRC
06:50:38 *** jmd has joined #gnucash
06:52:08 *** floh1111 has joined #gnucash
07:09:06 *** jmd has quit IRC
07:55:53 *** ErKa has joined #gnucash
08:02:26 *** floh1111_ has joined #gnucash
08:07:11 *** floh1111 has quit IRC
08:15:12 *** jmd has joined #gnucash
08:17:39 *** floh1111_ has quit IRC
08:23:32 *** jmd has quit IRC
08:36:19 *** benoitg has joined #gnucash
08:49:49 *** ErKa has quit IRC
09:24:26 <lmat> I forgot, what is the trajectory for changing gnucash to c++ ? I remember the GUID transformation, but there were other things...
09:28:38 *** ErKa has joined #gnucash
09:32:48 <warlord> lmat: eventually everything..
09:32:59 <warlord> I think gnc_numeric is another early choice
09:33:05 <warlord> but jralls will have to answer you.
09:36:00 <lmat> Hey, warlord 's back!
09:36:31 <lmat> I was looking at g_hash_table_* but saw that it's a glib function, so it's probably trivial (it would be easy, and wouldn't buy us anything)
09:37:00 <warlord> yep. Had a wonderful holiday
09:37:30 <warlord> yeah, anything g_ is glib. eventually we'll rewrite code to not use that, but... not yet.
09:39:30 <lmat> warlord: yeah, it seems that it would be silly to change g_hash_table_insert to gnu_hash_table_insert... But what *would* be a fun switch is when everything that uses g_hash_table_insert is in c++, so that we can simply use std::map.
09:39:34 <lmat> without a wrapper
09:39:36 <lmat> But yeah, not yet.
09:42:42 *** MechtiIde has joined #gnucash
09:43:13 <lmat> I'm looking at gnc-numeric. It has a int main;
09:43:18 <lmat> This should probably be moved to a test, right ?
09:43:20 <lmat> warlord: you can do that ?
09:43:38 <warlord> lmat: isn't that "main" inside an #ifdef?
09:43:57 <lmat> oh yeah, it is. Definitely needs to be moved to a test, then ^_^
09:47:13 <warlord> Depends -- it's possible the test code just recompiles the existing source with the #define set.
09:47:23 <warlord> So, there may not be a need to "move" the code, per se.
09:47:30 <lmat> hmm
09:55:51 *** jralls has quit IRC
09:56:23 *** jralls has joined #gnucash
09:56:23 *** gncbot sets mode: +o jralls
09:59:12 <lmat> Where is the code for a specific backend?
09:59:36 <jralls> lmat: The testing for gnc-numeric is a bit weird. It's over in src/engine/test, probably because it didn't get moved to qof when gnc-numeric itself did.
09:59:52 <lmat> I see
10:00:47 <jralls> Code for specific backends is in src/backends; they implement an interface in src/libqof/qof/qofbackend.
10:00:53 <lmat> excellent
10:01:22 <jralls> It's tightly coupled with qofsession, so you need to understand that code as well as the others.
10:03:25 *** isaacd has joined #gnucash
10:10:44 <jralls> Oh, and QofBackend obviously participates in QofInstance's edit/commit cycle, which is used throughout the engine, so you'll want to study that as well.
10:24:10 <lmat> jralls: :)
10:26:25 <jralls> Back to gnc-numeric, remember the discussion about decimal floats from last week. I'm going to start on that test right after getting boost into Win32.
10:26:54 <lmat> jralls: On the mailing list ?
10:27:13 <jralls> That's where the discussion was, yes.
10:27:30 <lmat> jralls: Yes, I just read it in the last hour or so :) Sounds like a good idea.
10:28:10 *** Krzysiek_K has joined #gnucash
10:29:10 <jralls> I think it will be. I was looking at the library docs last night. It's multiprecision, so it should settle all of the overflow issues as well as the binary rounding.
10:30:40 <jralls> On the other hand, when working on a scheduled transaction bug earlier this week I discovered that the interest rate calculations, pmt, ipmt, and ppmt, are done in Scheme using regular floating point.
10:31:30 <lmat> Oh, that's fun! (@scheme)
10:31:53 <lmat> It's not arbitrary precision, right ? You had suggested a struct { int64; int64 }; is that still the maximum size ?
10:42:17 *** GabrieleV has quit IRC
10:46:03 <warlord> Yeah, all SX functions are implemented in scheme. I guess technically you could implement it in C and then export it as Scheme, but that would be rather.. odd...
10:48:52 <jralls> lmat: Storage will have to be some sort of fixed precision, but libmpdecimal is multiprecision internally. I haven't yet figured out how to wrap that.
10:50:39 <jralls> warlord: AFAICT everything in SX is C *except* the interest functions. But even if that's not the case, why would it need to be exposed to Scheme and why would it be odd to export C code to Scheme?
10:52:45 <warlord> jralls: it would be odd to expose it just so the SX code can find it.
10:53:01 <warlord> IIRC, all SX "functions" need to be scheme symbols.
10:53:49 *** GabrieleV has joined #gnucash
11:07:05 <jralls> I think you mis-remember. See the Doxygen docs for src/app-utils/calculation/expression_parser.c.
11:17:38 *** MechtiIde has quit IRC
11:25:06 *** MechtiIde has joined #gnucash
11:29:50 <jralls> lmat, gjanssens: Here's a show-stopper: The MinGW 11 version of gcc doesn't support C++11. I'll proceed with Gnucash-on-windows, but I can't merge it back into main unless we're ready to switch the build server to use gnucash-on-windows.
11:36:15 *** MechtiIde has quit IRC
11:40:11 *** aqua___ has joined #gnucash
11:46:23 <lmat> jralls: Sorry, I don't know what you mean by "I'll proceed with Gnucash-on-windows". Does that mean you'll continue merging c++11 features? Or you'll continue without c++ on Windows? Or something else ?
11:47:44 <warlord> lmat: the gnucash-on-windows is a "new" windows build server substrate (IIRC)
11:47:51 <warlord> the current build server doesnt use it, yet.
11:47:54 <jralls> Something else: Gnucash-on-windows is an alternate set of build scripts Geert wrote using the current MinGW environment. You can get it from https://github.com/Gnucash/gnucash-on-windows if you're interested.
11:48:12 <lmat> I see. Thanks
11:48:20 <lmat> How do I get to the "add commodity" windodw ?
11:48:37 <jralls> Tools>Security Editor.
11:49:04 <lmat> thanks!
11:49:23 <jralls> Or from the Account editor, click the "select commodity" then "New" in the picker dialog.
12:00:45 *** Bodhi-Baum has joined #gnucash
12:00:52 *** benoitg has quit IRC
12:07:30 *** jmd has joined #gnucash
12:23:38 *** MechtiIde has joined #gnucash
12:28:12 <lmat> How can I add debugging statements in gnucash? I stuck a g_print("hey"); in there, but it isn't printing (and I'm pretty sure the code was called ^_^)
12:28:46 <warlord> lmat: have you looked in the gnucash.trace for your "hey"?
12:29:00 <lmat> warlord: nope! I was looking at std out.
12:29:15 <lmat> warlord: where should that be ?
12:29:53 <warlord> probably /tmp (at least on Linux)
12:30:06 <lmat> awesome, there it is!
12:30:29 <lmat> ahh, yup, there are the errors, too! (concerning why what I added didn't work :))
12:30:37 <warlord> :)
12:36:13 *** isaacd has quit IRC
12:37:48 *** jmd has quit IRC
12:38:28 *** jmd has joined #gnucash
12:48:36 *** Bodhi-Baum has quit IRC
12:48:48 *** Bodhi-Baum has joined #gnucash
13:12:30 *** aqua___ has quit IRC
13:27:15 *** Krzysiek_K has left #gnucash
13:29:57 *** Bodhi-Baum has quit IRC
14:16:34 *** aqua___ has joined #gnucash
14:21:31 *** floh1111 has joined #gnucash
14:31:06 *** aqua___ has quit IRC
14:32:01 *** MechtiIde has quit IRC
14:35:31 *** floh1111 has quit IRC
14:58:19 *** StuM1 has quit IRC
14:59:03 *** StuM has joined #gnucash
15:06:51 *** jmd has quit IRC
15:24:19 *** rpg has joined #gnucash
16:10:45 <gjanssens> jralls, warlord: the build server *is* running on gnucash-on-windows
16:11:05 <gjanssens> I have personally taken care of that a while back
16:11:20 <jralls> Oh, cool. Can we get rid of the old instructions in the wiki, then?
16:11:25 <gjanssens> It builds both master (nightly) and maint (weekly) from the gnucash-on-windows scripts
16:11:37 <gjanssens> jralls: yes
16:11:47 <gjanssens> I forgot to clean that part up.
16:11:55 <gjanssens> Feel free to do so if you like
16:12:03 <gjanssens> I'm ready for bed ;)
16:12:32 <gjanssens> The most recent usage instructions can be found in the Readme document(s) in the gnucash-on-windows repo
16:13:27 <jralls> OK. I'm still struggling with getting boost to build at all on MinGW.
16:15:20 <gjanssens> BTW the packaging/win32 directory is still in the gnucash repository because I still wanted to double check for hard-coded dependencies to there
16:15:31 <gjanssens> But I got side tracked by my dayjob
16:15:50 <gjanssens> I'm mentioning it here as a mental note to myself.
16:15:58 <gjanssens> See you later
16:16:05 <jralls> What do you mean by hardcoded dependencies?
16:16:11 *** gjanssens has quit IRC
16:26:29 *** fell_ has quit IRC
16:37:36 *** rpg has quit IRC
16:45:34 *** rpg has joined #gnucash
17:18:10 *** jralls has quit IRC
17:18:51 *** jralls has joined #gnucash
17:18:52 *** gncbot sets mode: +o jralls
17:35:52 *** jralls has quit IRC
17:36:46 *** jralls has joined #gnucash
17:36:46 *** gncbot sets mode: +o jralls
18:00:08 *** StuM has quit IRC
18:00:51 *** StuM has joined #gnucash
18:01:10 *** rpg has quit IRC
18:34:29 *** jralls has quit IRC
18:35:21 *** jralls has joined #gnucash
18:35:21 *** gncbot sets mode: +o jralls
19:29:22 *** isaacd has joined #gnucash
20:03:47 *** ErKa has quit IRC
21:14:09 *** ErKa has joined #gnucash
21:40:57 *** linas__ has joined #gnucash
21:40:58 *** linas_ has quit IRC
21:58:37 *** isaacd has quit IRC
23:22:43 *** ErKa has quit IRC
23:53:40 *** MechtiIde has joined #gnucash