2007-02-03 GnuCash IRC logs

01:49:09 *** jpeach has joined #gnucash
02:08:48 *** warlord is now known as warlord-afk
03:03:16 <BC-bd> *yawn*
03:03:32 * foo yawns with you
03:56:30 *** rlaager has joined #gnucash
04:18:40 *** bonez39 has quit IRC
04:46:23 *** Esaj has joined #gnucash
05:25:05 *** rlaager has quit IRC
05:26:04 *** rlaager has joined #gnucash
06:58:24 *** conrad has joined #gnucash
07:05:07 *** andi5 has joined #gnucash
07:05:07 *** gncbot sets mode: +o andi5
08:10:31 *** serge has joined #gnucash
08:13:17 *** serge has quit IRC
08:14:11 *** gander has joined #gnucash
08:46:50 *** |gunni| has joined #gnucash
08:50:50 *** mnoir has joined #gnucash
08:51:24 *** ErKa has joined #gnucash
09:03:20 *** ErKa has quit IRC
09:16:10 *** besh has joined #gnucash
09:28:35 *** besh has quit IRC
09:33:32 *** besh has joined #gnucash
10:16:17 *** vodnik has joined #gnucash
10:19:30 <vodnik> hi, I installed gnucash to managa my modest incomes&expenses and it seems that account equity (that hould actually show how much is mine) does not work, actually it only shows initial balances, but when I put something to income or expenses, equity does not change... what can be the problem? assets change as they should...
10:19:47 *** warlord-afk is now known as warlord
10:19:53 <warlord> vodnik: run the Balance Sheet report.
10:20:19 <jsled> vodnik: There's two concepts: the Equity *account* and *equity*, which you need to run a report to generate.
10:20:21 <warlord> gnucash is working as expected; the Equity accounts show you the equity the last time you "closed the books"
10:22:01 <vodnik> I see... well I am new to all this accounting stuff so I am rather confused, but Balance Sheet report works well, thanks
10:24:37 <vodnik> bye
10:24:50 *** vodnik has left #gnucash
10:28:23 *** besh has quit IRC
10:29:30 *** besh has joined #gnucash
10:31:36 *** besh has quit IRC
10:38:16 *** besh has joined #gnucash
10:47:12 *** besh has joined #gnucash
11:04:39 <andi5> hiho.... are there any objections to http://pastebin.ca/raw/338608 ? .... %x is used already in the code (the example locale date string looks different from the one actually used in the register and date entries), %x, %c and %X are C89 strftime format specifiers and supported by windows :) ... jsled?
11:05:35 <warlord> andi5: does reset.sh still properly remove everything that needs to be removed in order to rebuild gnucash?
11:06:03 <andi5> warlord: i have never used that command... i rarely trust automatic removal scripts
11:06:51 <andi5> but i guess it will miss at least inno, aqbanking, gwenhywfar
11:07:12 <warlord> I dont think I've built since those were added.
11:07:37 <warlord> We changed the openssl version, didn't we?
11:07:45 <andi5> yes
11:07:54 <andi5> but install.sh should complain about some peculiarities
11:08:37 <jsled> andi5: huh... [[[
11:08:43 <jsled> >>> time.strftime('%Y-%m-%d', time.localtime())
11:08:43 <jsled> '2007-02-03'
11:08:43 <jsled> >>> time.strftime('%x', time.localtime())
11:08:43 <jsled> '02/03/07'
11:08:43 <jsled> >>> time.strftime('%X', time.localtime())
11:08:44 <jsled> '11:13:27'
11:08:46 <jsled> >>> time.strftime('%c', time.localtime())
11:08:50 <jsled> 'Sat Feb 3 11:13:30 2007'
11:08:52 <jsled> ]]]
11:09:26 <andi5> what... is wrong about that?
11:10:39 <jsled> Well, I like %Y-%m-%d better than %x. :) And %c is just atrocious.
11:11:02 <andi5> jsled: but you trash german dates on windows by that
11:11:30 <jsled> How?
11:11:43 * andi5 still needs to check where we use qof_print_date_time_buff ...
11:11:49 <warlord> ==jsles
11:12:20 <andi5> well, there is no langinfo.h on windows... so the register and date entries print 2007-02-03, where the example locale string ("%x") in the preferences dialog shows 03.02.2007
11:13:05 <warlord> I thought the whole point of the preference was to control the string passed to strftime()?
11:14:24 <andi5> no, we just use qof_date_format_set(QofDateFormat df), were df might be QOF_DATE_FORMAT_LOCALE
11:15:00 *** mnoir has quit IRC
11:15:02 <warlord> That's fine, but the preference dialog should match.
11:15:14 <andi5> jsled: qof_print_date_time_buff is not in use anywhere, so the same will hold for %c
11:15:25 <jsled> I'm confused.
11:15:51 *** jpeach has left #gnucash
11:16:22 <andi5> http://svn.gnucash.org/trac/browser/gnucash/trunk/src/gnome-utils/dialog-preferences.c#L1382
11:16:50 <jsled> Oh, indeed.
11:16:58 <jsled> Well, that's very silly.
11:18:30 <andi5> actually, i do not understand why %x / %X / %c has not been used..... imho that is _the_ way to print and parse a date/time/datetime in locale format
11:18:41 <jsled> I was just thinking the same thing.
11:18:54 <jsled> Are they a "modern" invention?
11:19:31 <andi5> see c89... i just checked my german kernighan&ritchie (hey, my dad really bought a good book :))
11:20:07 <andi5> oh, i may also say that it solves an issue for me ;-)
11:22:08 <andi5> warlord: there is a gconfd-2.exe console opening again, but i am working on it .... will keep you informed
11:22:31 <warlord> andi5: okay. I'm just rebuilding to re-test 'make check'
11:22:58 <jsled> Yeah, it seems fine, I think. If the user wants a different format, so be it. But "%{x,X,c}" seem better than our own, and the prefs dialog should *definitely* match! In fact, why is is not using the #define?!
11:23:41 <andi5> well, the macro is defined in gnc-date._c_ ...
11:25:18 <andi5> jsled: i do not want to refactor gnc-date, that is too much work at the wrong time :)
11:25:56 <hampton> Hold on. What problem are you solving with these proposed changes.
11:26:34 * hampton is trying to come up to speed
11:26:59 <andi5> hampton: http://lists.gnucash.org/logs/2007-02-03.html#T11:12:20
11:32:15 <andi5> damn, my date edit crashed, but i cannot move it away as it got fixed by gdb ;-)
11:32:34 <andi5> it is in the way!!!
11:33:51 <warlord> haha
11:34:14 <warlord> (currently my win32 build is at guile)
11:34:45 <hampton> After looking at the code, I'm ok with the change. I have a question though. Why does FreqSpec.c always use the locale date format instead of respecting the preference setting?
11:35:56 <andi5> good question.... jsled? :)
11:36:49 <hampton> and afaict, qof_print_date_time_buff() isn't used anywhere, but that's not the only place GNC_D_FMT is used. Its used by qof_date_text_format_get_string() which is called by gnc_date_format_refresh() which is in the widget used to display dates.
11:36:49 <warlord> [OT] http://www.degredo.net/
11:37:38 <warlord> I think qof_print_date_time_buff() was an API I was going to use for invoice dates at one point, but never got that far.
11:37:50 <hampton> [ot] LOL. I love it.
11:38:52 <andi5> *g*
11:39:00 <jsled> http://www.penny-arcade.com/comic/2007/02/02
11:40:24 <warlord> Grr...
11:40:25 <warlord> "You have uninstalled the wrong version of OpenSSL, but its DLLs libssl32.dll, libeay32.dll, and ssleay32.dll are still existing in $WINDIR\\system32. You have to delete (or rename) them manually. However, if you know these DLLs are needed by some other package, please contact the gnucash authors so that we can adapt this script.
11:40:27 <jsled> hampton: it just seems to use it in one place (formatting the string "Once: ${date}") ... but, yes, it's wrong.
11:40:59 <andi5> warlord: yeah, did it complain clearly enough? :)
11:41:14 <jsled> chris: around?
11:41:54 <warlord> I suppose.. But still.. :(
11:42:20 <andi5> the good thing is that you probably only lost 4 minutes or so
11:42:47 <warlord> True.. Hopefully the new openssl installation will overwrite /bin/openssl.. which was a very old version.
11:43:31 <andi5> which old version do you mean? and what new? *confused*
11:43:54 <warlord> Old: whatever was installed before (no idea)
11:44:12 <warlord> new: whatever it's installing now.. (0.9.7c)
11:44:25 <andi5> well, it was newer than 0.9.7, because we installed it
11:44:42 <andi5> the installer put everything into system32 ...
11:45:37 <warlord> Well, there's also a /bin/openssl which I think was a 0.9.6 version..
11:45:41 <warlord> er, /bin/openssl.exe
11:46:14 <chris> jsled: here
11:46:14 <gncbot> chris: Sent 1 day, 18 hours, and 17 minutes ago: <andi5> what is your opinion about http://lists.gnucash.org/logs/2007-01-31.html#T19:41:26 ? ... this poll was sponsered by gncbot
11:46:31 <andi5> ah... maybe we should mask that one?
11:46:37 <jsled> chris: hey. Just looking at the intersection of Recurrence and FreqSpec, wanted to confirm a couple of things.
11:47:29 <jsled> chris: It appears that "semi-monthly (15th + last day of month)" isn't really supported by Recurrence? As well, the multiple days-of-week.
11:47:47 <jsled> Since they're both done as "composite" FSs, which Recurrence doens't have.
11:49:48 <chris> Not be Recurrence, per-se, but RecurrenceList is treated as composition.
11:50:17 <jsled> Oh. /me headslaps
11:51:21 <jsled> awesome.
11:51:34 <chris> I suppose to be really general, you'd have to use only the recurrenceList api, but most of the time it would be a list of one.
11:51:41 * jsled nods
11:52:15 <chris> warning: the recurrenceList code didn't get tons of testing, because I didn't use it in budgets.
11:52:54 <jsled> So the only real questions left are whether to version the datafile or try to retain compat, and (related) whether to expose the nifty new Recurrences that are supported.
11:53:16 <jsled> chris: noted.
11:53:24 <warlord> does test-recurrence test the list api?
11:53:50 <jsled> Part of me wants to just version the datafile and have some sort of python script (or just xslt if we can get away with it) to handle backwards compat of the datafile.
11:54:04 <chris> warlord: I don't think so.
11:54:16 * warlord nods
11:54:50 <warlord> jsled: How about a File -> Export As GNC-2.0 ?
11:56:22 <chris> jsled: I would even go so far as to say that, with an go-to-old-version external script, as long as the data loss was explicit and reasonably limited, it could be lossy.
11:57:57 <jsled> warlord: Thought about it ... sounds like a bunch of new code I don't really want to add. :)
11:58:30 <warlord> jsled: well, we already (sort of) have the code..
11:58:59 <jsled> Another option is to disallow the "extended" Recurrence types and just keep writing out FreqSpec XML...
12:00:11 <jsled> Another option is to allow the "extended" Recurrence types, then either save FreqSpecs if possible or Recurrences if not, and tell the user their datafile won't be readable by 1.8/2.0, and/or offer the "Export as Gnc-2.0"
12:00:15 <jsled> Bleh.
12:02:13 <chris> [ot] some runaway beagle cron job brought my system to its knees. :(
12:04:55 <hampton> [ot] something seems to be thrashing my system too, although my load is only 0.6
12:11:15 <andi5> ok... so my patch will not work... problem -> strptime is missing and lib/libc/strptime cannot handle %x & friends.... /me retries
12:18:29 <warlord> Heheh
12:30:27 *** warlord2 has joined #gnucash
12:30:28 *** gncbot sets mode: +o warlord2
12:30:52 *** warlord has quit IRC
12:35:52 <hampton> bbl
12:35:58 *** hampton is now known as hampton|away
12:43:57 <rlaager> andi5: Yes, %x, %c, and %X are supported on Windows. %z is not, and %Z returns significantly different values.
12:43:59 *** ErKa has joined #gnucash
12:45:43 <rlaager> andi5: I've been wondering if there shouldn't be a library (or functions in glib or GTK+) to abstract the difference, and to take into account the user's date format settings in GNOME/KDE instead of glibc, which you can't easily change.
12:46:07 <andi5> rlaager: are you talking about strftime or strptime?
12:46:47 <rlaager> andi5: If that's a question between the two, I'm talking about strftime. If you're saying that strftime() does what I want, then no, it doesn't. ;)
12:46:52 <andi5> rlaager: and what date format settings in GNOME do you mean? are there some?
12:47:22 <rlaager> We have a similar issue in Gaim in that some people really want their locale format, and some really don't. So we, like GnuCash, have a pref.
12:47:34 <rlaager> Aren't there? I could've sworn there were. I think KDE has some.
12:48:04 <andi5> i use kde every three years or so... and cannot remember such a setting gnome actually
12:49:09 <rlaager> Hmm, well, if there isn't one, maybe there should be, to eliminate app-specific prefs.
12:50:20 <andi5> rlaager: well, maybe it is time for a bug :)
12:50:27 <andi5> report, i mean
12:53:07 <andi5> rlaager: but i am unsure... how many levels of user configuration should exist? locale, desktop, application, ... what else? ... people can never get enough :( ... so the major question is: those that demand application specific date formats, do they want the same format for all desktop applications or do they want a different date on transactions than gaim logs?
12:54:02 <rlaager> andi5: That's a good question. In our case it's actually a big deal for time, 12-hour vs. 24-hour. You seem to be talking more about dates?
12:54:36 <andi5> yes, but we have some setting for 24hours too.... will have to look where we actually use it :)
13:02:21 <andi5> rlaager: so that is what i expected, gaim does not make use of strptime at all :) ... for gnucash is the situation is a bit different :(
13:14:52 <andi5> warlord2: what does your gnucash build say?
13:15:50 <warlord2> andi5: I dont know. Let me check
13:15:51 *** warlord2 is now known as warlord
13:15:52 <warlord> @op
13:15:53 *** gncbot sets mode: +o warlord
13:18:28 *** warlord2 has joined #gnucash
13:18:29 *** gncbot sets mode: +o warlord2
13:24:16 <warlord2> andi5: it's building aqbanking.
13:27:22 *** warlord has quit IRC
13:30:56 <andi5> warlord2: ok, so it will take some time to complete :) .... /me departs
13:31:17 *** andi5 has quit IRC
13:35:24 *** warlord2 is now known as warlord
13:50:04 *** nomeata has joined #gnucash
13:58:05 *** ErKa has quit IRC
14:10:09 *** sjc has joined #gnucash
14:15:09 *** jpeach has joined #gnucash
14:15:17 *** jpeach has left #gnucash
14:29:57 *** TheDude has joined #gnucash
14:41:08 <TheDude> I have a question regarding gnucash 2.0.x (I'm using 2.0.4 currently, but this problem existed in the 2.0.x series). I have imported some data from MS Money, but now when I open gnucash, it doesn't let me resize the window to fit my screen. I can only run 1024x768. The height is fine, but the width will not fit on the screen. Mostly it's the scroll bar and about half of the down arrow in...
14:41:09 <TheDude> ...the grid header. I tried finding if there's a minimum size limit in gnucash, but didn't find any requirements. The main account screen comes up with the "Account Name" and "Total" headers. These columns only take up about half the screen, the rest is white space. Any help will be greatly appreciated. Thanks.
14:43:23 <warlord> Yeah, there appear to be some issues with down-sizing the window in some instances. Try filing a bug with Bugzilla?
14:45:30 <TheDude> I haven't yet, but I can fill one out. At least now I know it's a regular bug issue rather than an "import from MS Money" issue. Thanks.
14:48:45 *** TheDude has left #gnucash
14:51:03 *** mnoir has joined #gnucash
14:51:46 *** mnoir has left #gnucash
15:08:40 <warlord> Yeah, it has nothing to do with MS Money
15:43:46 <foo> Can gnucash create a profit and lost statement ?
15:43:49 <foo> loss
15:44:29 <chris> foo: yes, it's called a "Profit and Loss Report", naturally enough.
15:44:48 <foo> chris: hm, where at? I couldn't find it in repors
15:44:50 * foo checks again
15:45:35 <foo> chris: Yeah, under what? I can't seem to find it
15:47:30 <chris> foo: ahh, it was renamed in 2.0...
15:47:44 <chris> let me try to remember what to...
15:47:57 <foo> ah, ok
15:48:08 <chris> Income statement
15:48:51 <chris> http://wiki.gnucash.org/wiki/FAQ#Q:_What_happened_to_my_Profit_and_Loss_report_when_I_upgraded_to_2.0.3F
15:49:02 <foo> ah
15:56:02 <foo> chris: Thanks man
15:56:17 <chris> foo: yw
15:56:56 <prock> jsled: if I create multiple GncSxInstanceModels will _gnc_sx_instance_event_handler be called for each one on every SX add/remove/edit?
16:08:41 *** Wilddev has joined #gnucash
16:08:41 *** gncbot sets mode: +o Wilddev
16:09:05 <Wilddev> hiya
16:11:49 <chris> hey Wilddev, long time no irc.
16:12:20 <Wilddev> chris: yeah been really busy with work
16:12:44 <chris> ah, same here.
16:13:52 <Wilddev> hmm 214 msgs in gnucash-devel, guess it has been a while
16:19:47 <jsled> prock: yup. Each one registers it's own qof event handler.
16:19:55 <prock> oh I was just writing a reply to your email
16:20:42 <prock> when you say "SX List: wants disabled SXes in the model, but not projected." ... by 'projected' are you meaning projected like the projection report? or that you don't want the actual instances created, just the model..?
16:20:52 *** MrN has joined #gnucash
16:21:24 <jsled> yes. :)
16:21:30 <prock> :p
16:21:46 <jsled> I was thinking, the projection report probably wants the instance model to actually model the instances that'd be created for the disabled SXes.
16:22:09 <MrN> hi
16:22:14 <jsled> But the SX List doesn't care for the instances ... or even the "next occurred" date for disabled SXes.
16:22:28 <jsled> It just needs to know that there *is* a (disabled) SX in the model.
16:22:40 <jsled> (To map to an entry in the list/model).
16:23:34 <prock> re: projection report: it sounds as though you would expect that a report would update automatically if I removed/edited a SX, is this what you were thinking?
16:24:01 <jsled> prock: not really; the reports don't exactly work that way
16:24:12 <prock> I didn't think so.
16:24:14 <jsled> (though it would be cool.)
16:24:25 <jsled> I mean, that's just a side effect of the way the reports are done now.
16:24:47 <jsled> But in an ideal case, the reports would be "live", and listen for events/signals, and update themselves appropriately.
16:24:48 <prock> In my mind, I only see us needing the one gboolean: include_disabled
16:25:00 <prock> I'm not sure I'd want my report to do that.
16:25:30 <prock> by "my" I mean any report... I don't want it to change unless I've explicitly told it to re-run.
16:25:35 <jsled> Most of the reports only regenerate when 'Reload' is explicitly called; that's fine.
16:26:28 <jsled> I had only 'include_disabled' in an earlier draft ... but then I realized that the report itself will want the disabled sx instances to be generated.
16:26:51 <jsled> I guess we could say that *any* SXes included in a model always have instances generated.
16:26:59 <jsled> Which simplifies things a bit.
16:27:26 <prock> re: gboolean's for gnc_sx_instance_model_new - SX List: wants to know about everything, Calendar - only enabled, SLR - only enabled, SX Report Dialog - everything. The report itself shouldn't be keeping an instanceModel around once it's been generated should it?
16:28:02 <jsled> Except that complicates the calendar model, since something will need to filter the instances of disabled SXes.
16:28:37 <jsled> prock: naw, the report can get and unref the model during its "work" phase; once it generates the report data, it doens't need the model around after that.
16:30:01 <prock> re calendar: how would it be complicated? The filtering would have already happened when it was first generated and any future filter would happen via _gnc_sx_instance_event_handler and include_disabled.
16:31:17 <jsled> For the SX List case: include_disabled=true, and instances are generated. Thus, there are "phantom" instances for the disabled SXes that the calendar would either a) display b) need to filter.
16:31:35 <prock> re sx projection getting events - I don't think anything but the SX Projection dialog (where the SX's are selected) would need (or want) to receive events. Once I select my events I want the report generated with the SX's just as I selected them.
16:31:50 <jsled> prock: sounds fine.
16:32:15 <prock> ah, so the sx list is sharing the instancemodel with the calendar?
16:32:53 <jsled> Yes, in that case. That is, the gnc_plugin_page_sx_list gives the same instance-model to both
16:33:11 <prock> would you be opposed to the calendar having it's own?
16:33:12 <jsled> (via the relevant adapters, of course)
16:33:13 <jsled> Yes.
16:33:48 <jsled> I mean, that's the point, right? They're both views of the *same* model.
16:34:18 <prock> The list and the calendar are not the same model: the calendar should only show enabled instances whereas the list would have to show everything
16:35:36 <jsled> Well right now, they are the same model. Logically, they are the same model. You're suggesting that they become different, but that doesn't seem right to me.
16:36:26 <prock> If they both use the same model I don't see how we could keep non-enabled sx's from showing on the calendar
16:36:30 <jsled> The filtering could be readily in the dense-cal adapter.
16:37:02 <warlord> Wilddev: Hey, man! How goes it?
16:37:13 <prock> oh, I thought you said you didn't want that: http://lists.gnucash.org/logs/2007-02-03.html#T16:28:02
16:37:24 <jsled> The model could have that second 'project_disabled' or 'create_disabled_instances' flag ... though as you suggest, that's less than ideal.
16:38:01 <jsled> No, I just said it would complicate it. :)
16:38:12 <prock> Ah :)
16:38:55 <jsled> At some future point, the dense-cal should probably get the ability to display more than one mark-type...
16:39:11 <jsled> ...then it could just display a different mark for disabled instances (maybe via UI toggle) or something.
16:39:25 <jsled> But, for now, I think it just filters them. That seems fine.
16:39:32 <jsled> And it's not much of a complication anyways.
16:39:41 <prock> Ok, so here's what we've ended up with...
16:40:03 <prock> SX List - process all, calendar skips non-enabled SX's
16:40:15 <prock> SLR - only deal with enabled SX's
16:40:31 <prock> Report dialog - process everything
16:40:38 <prock> does that sound good?
16:40:57 <jsled> yeah, then we only need that one flag to the instance_model c'tor, which is much better. :)
16:41:08 * prock nods
16:42:07 <prock> do you want me to file a bug about the recurrence not getting set to "Never" and the incorrect confirmation dialog?
16:42:33 <Wilddev> warlord: sorry was reading email, I'm good thx
16:42:41 <Wilddev> how ru?
16:42:51 <jsled> Um. Probably not. I've not found time to look at it today, but I probably won't forget it. I might add a note to sx.rst.
16:42:52 <warlord> Not too bad.
16:43:11 <prock> ok, I'll go look at _gnc_sx_instance_event_handler
16:43:13 <warlord> (and I know you've got a lot of email waiting)
16:43:21 <Wilddev> lol
16:43:36 <Wilddev> life has been kinda hectic lately
16:43:47 <warlord> Tell me about it!
16:46:39 <prock> jsled: one thing about the SX projection dialog - if I don't update the model that the report generator is using and while the report is being generated a SX is removed is there a potential segfault? The report generator's pointer to the SX would be pointed to free'd memory.
16:47:13 <jsled> prock: that's right. During the time you have a valid model, you need to respond to the events.
16:47:27 <prock> Hmmm... I don't think it's a show stopper right now but I'll keep it in mind
16:47:52 <jsled> "responding" to the events on the instance model is a bit broken, too...
16:48:54 <jsled> ... since the instance model totally violates encapsulation, it's the responsibility of the user to mutate the model to respond to the event; but the handlers are provided by the model, so it's just a matter of calling them.
17:06:29 <prock> jsled: why did you suggest that gnc_sx_instance_model_new also needs the range_end parameter, isn't that dealt with in gnc_sx_instance_model_new after the model is created?
17:07:56 <jsled> Well, I guess there is no gnc_sx_instance_model_new(...). I think I meant a modification of gnc_sx_get_instances(GDate *range_end).
17:08:36 <prock> there is static GncSxInstanceModel* gnc_sx_instance_model_new(void)
17:09:50 <jsled> prock: yeah, I didn't mean that one ... I was making up API. Sorry about that.
17:10:00 <prock> :)
17:10:09 <prock> Ok I get what you're saying
17:17:16 *** Esaj has quit IRC
17:30:10 <prock> jsled: will this strip out instances of 'sx' from 'instances'? - "g_signal_emit_by_name(instances, "removing", (gpointer)sx);"
17:31:07 <jsled> not until the receiver of "removing" calls gnc_sx_instance_model_remove_sx_instances
17:31:34 <prock> right but it does in effect do so. thanks.
17:33:09 *** benoitg has joined #gnucash
17:40:39 *** benoitg has left #gnucash
17:41:18 *** warlord2 has joined #gnucash
17:41:18 *** gncbot sets mode: +o warlord2
17:41:40 <prock> PERR is for logging problems, correct?
17:42:00 <warlord2> prock: yea.
17:42:06 <warlord2> PDEBUG, PWARN, PERR
17:42:10 <jsled> yeah; it's an error-level statement with printf syntax.
17:42:17 <prock> ok
17:42:18 <warlord2> I think we have a PINFO too.
17:42:30 <warlord2> It lets us control logging on a per "module" basis..
17:42:42 <warlord2> (so certain areas can be more verbose than others)
17:42:46 <prock> something like log4cplus
17:42:55 <jsled> yeah, something like.
17:43:02 <jsled> Though we should just use log4c
17:43:15 <jsled> but this has been around well before the log4X'es came into play, I think.
17:43:50 <prock> jsled: there's no reason why I can't use Pfoo in _gnc_sx_instance_event_handler, is there?
17:44:17 <jsled> prock: no, you should. I didn't just becuase I dislike the fact that Pmumbles don't go to stdout.
17:44:25 <jsled> but I should, too. :/
17:44:36 <warlord2> P*() goes to gnucash.trace
17:45:34 <prock> jsled: ah, I was confused why they weren't being used. I love those kind of logging macros
17:48:38 *** Wilddev has quit IRC
17:50:35 *** warlord has quit IRC
17:52:40 *** warlord2 is now known as warlord
18:07:01 *** mnoir has joined #gnucash
18:29:42 *** warlord is now known as warlord-afk
18:46:16 <conrad> jsled: (Assuming you're still the gncbot maintainer) Could you please update gncbot to also accept me with hosts 125.168.104.204 - 207?
18:48:59 *** gander has quit IRC
18:54:54 *** praetis has joined #gnucash
18:57:57 *** lewy has joined #gnucash
18:58:52 <praetis> how would I best record the expense split for a restaurant entry where I paid for the table with my card and others at the table gave me cash for their parts?
19:00:45 <lewy> GnuCash was working. Ran "port update outdated" (OK), then failed install of GRAMPS, Both under MacPorts. Now cash won't launch with "open-x11 gnucash". Attempt to launch as gnucash from within X11 terminal gets error:
19:00:49 <lewy> MacBook:~ lewy$gnucash
19:00:50 <lewy> dyld: Library not loaded: /opt/local/lib/libdbus-1.2.dylib
19:00:50 <lewy> Referenced from: /opt/local/lib/libgoffice-1.2.dylib
19:00:50 <lewy> Reason: image not found
19:00:51 <lewy> Trace/BPT trap
19:00:51 <lewy> MacBook:~ lewy$
19:01:10 <lewy> Board meeting next Wednesday. Eeek.
19:01:29 <lewy> Any help for the ill informed?
19:06:33 <jsled> lewy: hmm. I'm not a mac person, but that looks very similar toi the error I got when going from <dbus-0.91 to dbus-1.0.2
19:07:07 <lewy> sounds like a good guess. what did you do?
19:08:25 <jsled> lewy: rebuilt every package that has a (library) dependency on the now-missing lib.
19:08:43 <jsled> (In gentoo-ese: `revdep-rebuild`)
19:08:57 <jsled> Did you upgrade dbus recently? You might need to downgrade.
19:09:36 <lewy> probably. I ran an svn update outdated. Go back to dbus-0.91?
19:10:16 <jsled> lewy: Probably, or re-install/-build all the others.
19:10:20 <jsled> It's a binary package system, right?
19:10:32 <lewy> Do you know if there is an easy way to rebuild from svn?
19:10:42 <lewy> No, I think they are all compiled here.
19:11:09 <lewy> Good candidate for overnight on the Macbook. ;-)
19:11:14 <jsled> rebuild what from svn? gnucash? `./configure [options] && make install` should suffice.
19:12:23 <jsled> praetis: Well, the assets (or liabilities, depending on what kind of card it was) is fixed at whatever amount was actually charged. As well, *your* Expense:Dining is fixed (at your fraction of the meal)
19:12:26 <lewy> I was thinking more about backing out of dbus-1.0.2.
19:13:09 <jsled> lewy: I don't know how to do that in your distro/system; sorry. But it abstactly sounds like a reasonable course of action.
19:13:39 <jsled> praetis: the rest I'd probably just make an {Income:Misc -> Assets:Cash in Pocket} transation,
19:14:33 <lewy> I'll go try it. I'm installing with MacPorts, so I'm farther away from the process than the configure/make/install (as in there is a script).
19:14:37 <lewy> Thanks for help.
19:14:39 <jsled> conrad: you can admin gncbot via dialog.
19:15:33 <lewy> Hmmm. I said update via svn -- meant via MacPorts. Sorry.
19:16:08 <praetis> jsled: I've come to think I'll just add an extra transaction to my dining expense that puts money back in my wallet asset rather than adding another form of income
19:16:45 <jsled> praetis: except those funds need to source from somewhere...
19:17:06 *** lewy has quit IRC
19:17:31 <praetis> jsled: well they come out of my checking account to make the actual payment. Then I'm just pretending it's the restaurant giving me back the difference as cash when in fact it's the people with me at the restaurant.
19:18:44 <jsled> praetis: It might be the beer, but I don't see how that works ... what are the Splits in the transaction?
19:18:48 *** nomeata has quit IRC
19:19:42 <praetis> jsled: no splits. That's what I was hoping there might be, but I simply have one payment entry for the full charge on the card, and a second rebate entry for the cash I got back.
19:20:15 *** lewy has joined #gnucash
19:20:15 <jsled> There are always splits.
19:20:41 <praetis> I mean, no three-way fanciness.
19:21:12 <praetis> Assets:Checking -> Expenses:Restaurants; Expenses:Restaurants -> Assets:Wallet
19:21:16 <jsled> Okay, so you're sourcing the funds from the Expense... right.
19:21:22 <lewy> jsled: what was that library name? The 0.9.1 vs 1.0.1 that you mentioned. I mis-copied it.
19:21:50 <praetis> it's a distinction only worth considering within the context of the restaurant expenses. It'd be unwanted detail at the income level, I think
19:22:42 <jsled> lewy: for me it was "libdbus-1.so.2".
19:22:54 <jsled> but the package that provides it is just 'dbus'.
19:23:09 <lewy> thanks.
19:24:15 *** lewy has quit IRC
19:42:44 <rlaager> praetis: Alternatively, you could just merge those two transactions. So you'd have one split where you expense your portion of the bill, then one where you increase your wallet amount by the cash total, then the final split where you put the whole amount on your credit card. That should all balance... if not, either you paid more than your share, or you made a profit from the others. ;)
19:43:27 <praetis> rlaager: mm, that's what I was looking for. Thanks!
19:44:47 <rlaager> Since things like that never come out perfectly even, I usually just through the imbalance into Expenses:Adjustment, myself.
20:08:53 *** andi5 has joined #gnucash
20:08:53 *** gncbot sets mode: +o andi5
20:10:19 *** mnoir has quit IRC
20:24:54 *** jpeach has joined #gnucash
20:25:02 *** jpeach has left #gnucash
20:25:12 *** |gunni| has quit IRC
20:47:10 <prock> @tell jsled I cannot reproduce the incorrect confirmation dialog problem ... I think I may have been setting the recurrence frequency to "Weekly" and then not selecting any days of the week. However I still am unable to set the recurrence frequency to "Never" and have it persist.
20:47:10 <gncbot> prock: The operation succeeded.
20:54:28 *** mishehu has quit IRC
20:58:49 *** sjc has quit IRC
21:06:01 *** praetis has left #gnucash
21:44:07 *** motin has joined #gnucash
21:52:57 *** MisterN has joined #gnucash
22:02:29 *** MrN has quit IRC
22:20:28 *** MisterN has quit IRC
22:45:04 *** conrad has quit IRC
22:50:54 *** motin has quit IRC
23:29:33 *** andi5 has quit IRC
23:31:30 <foo> hm, ok, so I'm importing a qif from my bank .. and I have a transaction from office max and I bought CD-Rs, CD cases, a sharpie pack, and a mileage log... I don't know the best way to categorize this. I know I could put the different things I purchased in different expense categories... but, how can I do that with the importer? It seems I can only select 1 account for it to come from
23:33:53 <prock> foo: select an account and do a split later
23:34:26 <prock> foo: but it all sounds like "Expenses:Office Supplies" to me
23:35:22 <foo> prock: Hm, thanks for both.. would you consider CD-Rs office supplies, though?
23:36:04 <prock> foo: yes
23:36:26 <foo> hmm, ok.. thanks!
23:36:31 * foo does that
23:36:42 <foo> prock: hm, curious - where are you located?
23:37:03 <prock> in my office. my cd-r's on next to me
23:37:08 <foo> Hm, looks like this grouped all my PAYPAL withdraws... hmm, /me will split that later he supposes
23:37:10 *** jpeach has joined #gnucash
23:37:15 <foo> prock: haha. On the globe :)
23:37:15 *** jpeach has left #gnucash
23:37:39 <prock> canada
23:37:44 <foo> cool
23:40:12 <foo> so, when importing a qif .. it asks me to select an account name... that's where the money is going to? eg. if it was an office max thing, it would go for office supplies... and if it was a deposit, I would select checking account (assuming I'm importing a qif from my checking account), right?
23:40:52 <prock> sounds right to me
23:42:29 <foo> ok. Hm, I bought tickets for a conference. /me creates a group called Conference Tickets under Expenses