2007-02-25 GnuCash IRC logs

00:49:29 *** warlord has joined #gnucash
00:49:30 *** gncbot sets mode: +o warlord
00:53:42 <Wilddev> hey warlord
00:54:21 <warlord> hi..
00:54:37 * warlord is in Palo Alto, now.
00:54:43 <Wilddev> fun
00:55:20 <warlord> *nods*
00:56:34 <Wilddev> ok sleep time for me
00:56:36 <Wilddev> nite
00:56:59 *** Wilddev has quit IRC
01:11:02 *** conrad has joined #gnucash
01:38:51 *** warlord is now known as warlord-afk
05:01:00 *** IanL has joined #gnucash
05:45:22 *** andi5 has joined #gnucash
05:45:22 *** gncbot sets mode: +o andi5
06:02:45 *** kielein has joined #gnucash
06:56:56 *** gander has joined #gnucash
07:00:22 *** cortana has quit IRC
08:41:35 *** egsavage has joined #gnucash
08:45:37 *** |gunni| has joined #gnucash
08:58:31 *** IanL has quit IRC
09:17:16 *** hampton|away is now known as hampton
09:23:04 *** twunder has joined #gnucash
09:23:50 *** warlord-afk has quit IRC
09:46:13 <jsled> gncbot: tell warlord hopfully he can take a look at <http://bugzilla.gnome.org/show_bug.cgi?id=410153>.
09:46:13 <gncbot> jsled: The operation succeeded.
10:12:12 *** jpeach has joined #gnucash
10:12:39 *** jpeach has left #gnucash
10:18:35 <jsled> gncbot: tell warlord What's the use case for <http://bugzilla.gnome.org/show_bug.cgi?id=88181#c3>?
10:18:35 <gncbot> jsled: The operation succeeded.
10:32:24 *** warlord-afk has joined #gnucash
10:36:25 <jsled> andi5: around?
10:36:31 <andi5> yep
10:37:03 <jsled> andi5: are you otherwise modying the gnc-html code? I'm thinking of ripping out the form-submission handlers, and thus gnc-http which is all stubbed out with g_critical(...)s anyways.
10:37:29 <andi5> no... i was just fixing a crash that hit me
10:37:38 <jsled> noted.
10:42:10 <warlord-afk> *waves*
10:42:14 *** warlord-afk is now known as warlord
10:42:20 <warlord> hi
10:42:20 <gncbot> warlord: Sent 56 minutes ago: <jsled> hopfully he can take a look at <http://bugzilla.gnome.org/show_bug.cgi?id=410153>.
10:42:21 <gncbot> warlord: Sent 23 minutes ago: <jsled> What's the use case for <http://bugzilla.gnome.org/show_bug.cgi?id=88181#c3>?
10:47:48 *** jpeach has joined #gnucash
10:47:51 <warlord> jsled: w.r.t. 88181, I dont recall the actual use case off the top of my head, that was input four years ago. I think some of the reasoning was that I was doing payroll on the last business day of the month, and the transfer was happening two days before that, so I wanted to be able to say "3rd from last business day".
10:48:38 <jsled> Ah, it's 4-to-last business day, not 4th-to-last-weekday.
10:49:06 <jsled> like the 4th-to-last-monday, which might not even exist.
10:49:13 * jsled goes to get more coffee. :)
10:50:35 <warlord> I'm not sure I understand the distinction you're trying to make.
10:50:48 <jsled> 4 days vs. 4 weeks
10:51:34 <warlord> The other thing that would be useful is "N weekdays before the Nth day", eg 2 business days before the 15th
10:52:19 <jsled> yeah ... that and 'the previous/next friday/monday for a non-business day-of-month'.
10:52:27 *** Demitar_ is now known as Demitar
10:52:36 <warlord> *nods*
10:53:43 <hampton> can you get Easter in there too? You know, the first sunday after the first full moon after the vernal equinox? :-)
10:53:58 <warlord> as for 410153, chris_ should take a look at it. That's hitting the g_assert() on line 278 of Account.c asserting that all the Splits are out of the Account. There's no way that that crash happened while trying to delete a transaction.
10:54:13 <hampton> or is it "on or after the vernal equinox."
10:57:35 <warlord> jsled: did that answer your questions?
11:01:12 <jsled> warlord: yes, thank you.
11:01:30 <warlord> NP
11:15:12 <jsled> Hmm. Do we have a bug for the gkeyfile key-name issue?
11:15:24 <warlord> I dont know.
11:15:33 <warlord> @op
11:15:34 *** gncbot sets mode: +o warlord
11:16:03 <jsled> @op prock conrad benoit
11:16:03 <gncbot> jsled: Error: benoit is not in #gnucash.
11:16:06 <jsled> @op prock conrad benoitg
11:16:07 *** gncbot sets mode: +ooo prock conrad benoitg
11:16:49 <jsled> Cool; I'm able to remove src/network-utils entirely. :)
11:17:16 <warlord> Yay!
11:17:50 *** bmc has joined #gnucash
11:18:22 <hampton> warlord: I've got the gkeyfile changes waiting to commit on march 18th
11:18:41 <hampton> i.e. 2.0.5 release + 1 month
11:19:13 <jsled> why the 1m?
11:19:16 <warlord> okay....
11:19:45 <hampton> waiting for the strcmp->safe_strcmp fix to propagate
11:20:58 <warlord> It's mostly an issue for people who swap between 2.0 and trunk, right?
11:21:03 <hampton> right
11:21:22 *** jpeach has left #gnucash
11:21:44 <bmc> I've written a Perl script which will dump all (or parts) of a gnucash database to a csv file. I'd be happy to contribute it, if it would be useful.
11:22:20 <jsled> Cool. Might be a good thing to link to from the wiki.
11:25:48 <bmc> Is there a contrib section at the gnucash site for stuff like this?
11:26:03 <warlord> bmc: not really.. Just the wiki..
11:26:07 <warlord> Or the email list.
11:26:42 <bmc> I found it very useful for doing my taxes -- I wrote it a year ago just for that.
11:28:33 *** MrN has joined #gnucash
11:29:30 <warlord> How is this different than the HTML table you get from an export from the transaction report?
11:30:05 <bmc> Well, the csv can be input directly into a spreadsheet.
11:30:56 <MrN> hi
11:30:57 <bmc> Perhaps some spreadsheets can import html, but I'm not familiar with such a capability.
11:32:43 <warlord> I'm pretty sure that OOo and Excel can import an html table as a spreadsheet.
11:32:59 <bmc> I'll try that.
11:34:08 <andi5> Q: gnc-reverse-balance, exported in swig-app-utils.c does not seem to be re-exported anywhere... how do other modules make use of it then? .... my problem: it seems like _i_ need to re-export my symbol :-)
11:39:02 <warlord> andi5: Is the symbol not available to you?
11:39:08 <bmc> warlord, I can't import html as a table, but cut & paste works. Ugly, but usable. Ok for budget stuff, might overload the clipboard when there are many rows.
11:39:29 <warlord> hmm
11:43:46 <andi5> warlord: do not mind... it did not really change anything i have not tested before, but it works now...
11:44:08 <warlord> ok
11:45:48 *** andi5 has quit IRC
12:20:20 *** jpeach has joined #gnucash
12:24:49 *** jpeach has left #gnucash
12:38:47 *** bock has joined #gnucash
12:38:48 *** benoitg has quit IRC
12:38:57 <warlord> @op bock
12:38:58 *** gncbot sets mode: +o bock
12:39:11 *** bmc has left #gnucash
12:44:46 *** cortana has joined #gnucash
12:49:38 *** andi5 has joined #gnucash
12:49:39 *** gncbot sets mode: +o andi5
13:40:18 *** andi5 has quit IRC
13:44:10 *** andi5 has joined #gnucash
13:44:12 *** gncbot sets mode: +o andi5
13:46:27 *** cortana has quit IRC
13:52:10 *** bock has left #gnucash
13:53:56 *** cortana has joined #gnucash
13:55:42 *** benoitg has joined #gnucash
14:11:46 *** nomeata has joined #gnucash
14:53:11 *** motin has joined #gnucash
14:55:09 *** twunder has quit IRC
15:02:58 *** ErKa has joined #gnucash
15:20:30 *** cortana has quit IRC
15:33:32 *** blahdebla1 has joined #gnucash
15:34:27 <blahdebla1> Is it possible to group invoice payments together into a single transaction?
15:34:40 <warlord> blahdebla1: of course. gnucash will automatically do that.
15:34:55 <blahdebla1> How?
15:35:24 <warlord> Just enter in the total payment and gnucash will automatically map it invoices in a FIFO order.
15:35:51 <blahdebla1> So you don't have to select the invoice numbers first?
15:36:10 <warlord> Nope.
15:36:21 <warlord> It will apply them in FIFO order to a specific customer.
15:36:25 <warlord> Selecting the invoice number is optional.
15:36:32 <blahdebla1> Cool - thanks.
15:37:12 <blahdebla1> What about deleting invoices? I can't for the life of me find where there's an option to do that.
15:38:12 <warlord> blahdebla1: there isn't one. See the FAQ
15:39:34 <blahdebla1> OK - so i should just edit it for re-use when i need to?
15:41:46 <warlord> yes, or just mark it inactive and leave it.
15:41:58 <jsled> andi5 :)
15:43:37 <andi5> hampton: what time zone strings are can be expected as result? ... i have not yet seen one, only gnc:no-time-zone or so
15:43:47 <andi5> hampton: as result of a gnc-fq-helper request
15:50:06 <hampton> um, none?
15:50:50 *** Demitar has quit IRC
15:51:42 <andi5> ahem... none?
15:52:03 * hampton is looking at F::Q now...
15:53:39 <hampton> gnc-fq-helper certainly doesn't seem to ever expect to see a timezone
15:54:01 <andi5> hampton: btw, i tried to get a quote with europe, but it did not respond/return... this blocks the application now, did it do that before too?
15:54:25 <hampton> # gnc:time-no-zone is returned as a string of the form "YYYY-MM-DD
15:54:25 <hampton> # HH:MM:SS", basically the unmolested (and underspecified) output of
15:54:25 <hampton> # the quote source. It's up to you to know what it's proper timezone
15:54:25 <hampton> # really is.
15:54:26 <andi5> (maybe i did not wait long enough)
15:54:40 <hampton> from gnc-fq-helper
15:55:06 <hampton> I believe it always blocked the app.
15:55:27 <andi5> ok
15:56:12 <blahdebla1> Thanks for the answers - you guys are so helpful!
15:56:28 *** jpeach has joined #gnucash
15:56:29 *** blahdebla1 has left #gnucash
15:56:35 *** jpeach has left #gnucash
15:56:55 *** kielein has quit IRC
15:57:13 <andi5> hm... do we add that time zone info anywhere?
15:58:42 *** Demitar has joined #gnucash
15:59:02 <hampton> no, and I think you'll find a bugzilla about that somewhere. F::Q should be supplying the timezone information, not gnucash.
16:00:29 <andi5> well... i will just ignore time-zones for now.... someone will have to fix that later then :) (without going back to guiles mktime, of course)
16:03:08 <andi5> hampton: #339433?
16:03:56 <hampton> that's the one.
16:04:45 *** cortana has joined #gnucash
16:08:22 <warlord> fun
16:09:32 * hampton thinks gnucash should use GDate and not time_t for its internal dates
16:09:32 *** Demitar has quit IRC
16:09:51 *** Demitar has joined #gnucash
16:09:51 <jsled> unless we need times.
16:11:17 <hampton> The problem with time_t is all conversions are sensitive to timezones
16:11:59 <hampton> are you planning to add time-of-day support next? :-)
16:12:06 <warlord> does GTime have an explicit TZ?
16:12:29 <andi5> GTime is deprecated and is basically 32bits of time_t
16:12:34 <hampton> Simply a replacement for time_t. It has been deprected since it is not equivalent to time_t on 64-bit platforms with a 64-bit time_t.
16:12:51 <warlord> Oh.
16:13:03 <jsled> There's GTimeVal, which is just seconds + µseconds
16:13:55 <warlord> Oh, sorry.. I mean GDate, not GTime
16:13:56 <andi5> jsled: do you use that?
16:14:24 <jsled> No, not that I recall, anyways.
16:25:51 *** gander has quit IRC
16:33:58 *** gander has joined #gnucash
16:36:54 <chris_> jsled: I just commented on 410153.
16:42:43 <andi5> so what about : (1) rename lib/libc/strptime.h to libc-time.h (or so, someone has complained) ; (2) move src/backend/file/sixtp-utils : gnc-timegm to that file as timegm ; (3) add src/app-utils/guile-util : gnc-parse-time-as-gmt ; (4) add the necessary seconds for the timezone (with) dst in price-quotes.scm once this information is available
16:47:12 <hampton> huh? 1) no. who complained. 2) no. gnc-timegm has nothing to do with libc. 3) not sure what this is for. 4) agreed. if F::Q ever provides timezones gnucash should use them when converting the times to gmt.
16:47:13 <chris_> andi5: how 'bout just writing price-quotes.c?
16:47:30 <andi5> chris_: a lot more work?
16:47:42 <hampton> chris_: I already tried asking him that.
16:48:01 <chris_> andi5: really?
16:48:25 * hampton has check printing mostly working in C
16:50:54 <andi5> hampton: (2) why does gnc-timegm have nothing to do with libc? who else provides timegm on linux? .... (3) that function will be needed to convert a date string to a time_t from within guile , (1) well, i do not want to clutter lib/libc, so i thought i could put both function into one file ; and (5) do we still need lib/libc/setenv.[ch]? ... i think we use g_(set|get)env everywhere (will need to double-check that)
16:51:50 *** gander has quit IRC
16:52:21 <chris_> andi5: delete all the parentheses and there's only about 25 lines of code there.
16:52:23 <hampton> name a libc function that provides a gnc-timegm.
16:53:28 <hampton> If you can't then gnc-timegm shouldn't go in lib/libc. The sole purpose of that directory is to provide libc functions that are missing on some machine or other.
16:54:14 <andi5> #ifdef HAVE_TIMEGM ; # define gnc_timegm timegm ; #endif .. + http://linux.about.com/library/cmd/blcmdl3_timegm.htm = make me wonder why "timegm" would be an incorrect answer :)
16:56:25 <hampton> gnc_timegm != timegm. gnc_ implies its a gnucash function.
16:56:35 <andi5> hampton, chris_: maybe i will do the conversion to c next week, i just hope it will not render it unusabe :)
16:56:37 <hampton> what's the problem? is windows missing the timegm function?
16:56:44 <andi5> YES
16:57:05 <hampton> then you should create a lib/lib/timegm.c file to implement it, and add a test in configure for timegm.
16:57:25 <andi5> hampton: it is all there :)
16:57:28 <hampton> lib/libc/timegm.c
16:57:37 <andi5> ok, not that file, you are right :)
17:00:55 *** chris_ is now known as chris
17:01:47 <hampton> chris: Tired of dragging that extra underscore around? ;-)
17:02:16 *** nomeata has quit IRC
17:03:15 <chris> It was totally throwing my center of gravity off.
17:18:04 *** andi5 has quit IRC
17:33:57 *** benoitg has left #gnucash
17:37:28 *** benoitg has joined #gnucash
17:40:01 <prock> are swig modules different and completely separate from gnucash modules?
17:40:59 <prock> i.e. I'm de-modulifying gnucash/engine and it calls scm_init_sw_engine_module. Is this related to gnucash modules or is this just part of tying gnucash to scheme?
17:41:57 <chris> the "swig" modules are "modules" in the guile sense of the term, not in the GnuCash sense.
17:42:17 <prock> ok, thanks.
17:42:33 <chris> np
17:48:03 <jsled> [ot] http://earthobservatory.nasa.gov/NaturalHazards/shownh.php3?img_id=14126
17:48:46 <hampton> cool pic
17:51:22 <prock> yuck, why is gncmod-backend-file unloading the engine?
17:52:45 <prock> wow, they've got a 4.74MB version of that sat. photo
17:53:19 <prock> hmmm I can see warlord and jsled's houses...
18:01:39 *** motin has quit IRC
18:06:51 <jsled> *un*loading? where?
18:07:32 <prock> src/backend/file/gncmod-backend-file.c line 57
18:08:13 <prock> it's when it's ending so it's not that big a deal. it's just not what I'd expect it to do.
18:08:42 <warlord> prock: all the modules should unload the dependencies when they themselves are unloaded.
18:08:56 <warlord> It's from back-in-the-day of guile modules..
18:09:17 <prock> actually unload them or merely decrease the ref count?
18:10:03 <warlord> Both. But if module B refs module A but never unrefs it, then A can never be unloaded.
18:10:44 <prock> ok, makes sense then.
18:13:56 <warlord> 'k
18:14:02 *** motin has joined #gnucash
18:21:53 *** ErKa has quit IRC
18:38:23 *** |gunni| has quit IRC
18:39:29 *** hampton2 has joined #gnucash
18:39:29 *** gncbot sets mode: +o hampton2
18:58:28 <warlord> okay, off to some friends. later..
18:58:30 *** warlord is now known as warlord-afk
19:22:22 *** MrN has quit IRC
19:56:31 *** jpeach has joined #gnucash
19:57:31 *** jpeach has left #gnucash
20:08:52 *** twunder has joined #gnucash
20:22:48 *** gncbot sets mode: +o benoitg
20:55:50 *** prock_ has joined #gnucash
21:04:40 *** prock has quit IRC
21:41:22 *** twunder has quit IRC
23:08:53 *** jpeach has joined #gnucash
23:09:01 *** jpeach has left #gnucash
23:14:52 *** jpeac1 has joined #gnucash
23:14:59 *** jpeac1 has left #gnucash
23:27:03 *** jpeach has joined #gnucash
23:27:07 *** jpeach has left #gnucash
23:35:18 *** jpeac1 has joined #gnucash
23:35:22 *** jpeac1 has left #gnucash
23:39:28 *** egsavage has left #gnucash