2007-04-07 GnuCash IRC logs

01:02:10 *** Wilddev has joined #gnucash
01:02:10 *** gncbot sets mode: +o Wilddev
01:25:48 *** Ardonik has quit IRC
01:33:13 *** Ardonik has joined #gnucash
01:42:10 *** kielein has joined #gnucash
02:40:28 *** Wilddev has quit IRC
03:04:58 *** nbinont has quit IRC
03:43:03 *** lasindi_ has quit IRC
03:44:32 *** lasindi_ has joined #gnucash
05:10:25 *** goibhniu has joined #gnucash
05:18:18 *** rhowe has joined #gnucash
05:24:54 *** lasindi_ is now known as lasindi
05:36:50 *** cortana has quit IRC
05:47:46 *** cortana has joined #gnucash
06:22:34 *** ErKa has joined #gnucash
07:05:14 *** ErKa has quit IRC
08:07:06 *** leolein has joined #gnucash
09:36:25 *** twunder has joined #gnucash
09:48:49 *** ErKa has joined #gnucash
09:49:38 *** warlord-afk is now known as warlord
09:55:00 *** Ardonik has quit IRC
10:01:29 *** twunder has quit IRC
10:02:28 *** Ardonik has joined #gnucash
10:28:34 *** twunder has joined #gnucash
10:52:48 *** ErKa has quit IRC
10:56:26 *** leolein has quit IRC
11:02:18 <warlord> Hmm.. the price scatterplot dies when reading from my transactions. (but works from the PriceDB). I wonder what's wrong with my datafile?
11:02:50 <warlord> Ahh, a Check&Repair fixed.
11:03:27 * hampton2 is trying to be nice in his response to AS
11:04:53 <jsled> huh?
11:05:27 <hampton2> [ot] he just comes across as suck a jerk in his email. me, me, me. "The world should revolve around what I want and don't want. Everyone should bow to my wishes"
11:05:28 <warlord> jsled: re: F::Q on -devel
11:06:18 <jsled> right. I just didn't read it that way. He just wants an "API".
11:06:33 <warlord> but we already have THREE APIs.
11:06:40 <hampton2> then why all the rant about not wanting to learn perl?
11:06:49 <warlord> the gnc-fq-helper API, the C API, and the scheme API.
11:08:15 <jsled> I don't see a rant.
11:08:35 <warlord> jsled: oh, it could across as a rant..
11:08:43 <jsled> I see "I don't want to install a bunch of stuff just to read numbers out of a file.". Which might be misguided, but hardly qualifies as a rant.
11:08:55 <hampton2> gnc-fq-check and gnc-fq-helper could be collapsed into a single executable, and a run time argument specified to allow changing the executable
11:09:25 <hampton2> Points a through e in his second email
11:09:40 <jsled> I'd not even do it as a runtime arg. I just say: here's the file it's going to call, and here's the s-expr format that it expects to be output.
11:10:40 <hampton2> doesn't want to write a F::Q module, doesn't want to learn perl, doesn't want to ... but wants others to do things for him.
11:13:27 <jsled> Hmm. I just see a simple question about what the interface is. But whatever.
11:14:28 <hampton2> That's what I thought of his first email. It was the second I thought was a rant.
11:15:05 <hampton2> but whatever. I agree to disagree. :-)
11:15:56 <warlord> jsled: imagine he was talking about the SX editor.... You'd start to take it personally. ;)
11:16:20 *** _gunni_ has joined #gnucash
11:19:08 <jsled> perhaps, but I try not to.
11:20:12 <warlord> We all try not to, but it still doesn't stop it from happening periodically.
11:26:13 <warlord> Okay, I think I'm going to go merge my gobject-engine-dev-warlord branch into trunk.
11:26:22 <hampton2> Yeah!
11:26:58 <warlord> I take it you like this idea, hampton ?
11:27:05 <jsled> heh
11:27:23 <hampton2> whatever gave you that idea? :-)
11:27:40 <warlord> Oh, I dont know..
11:28:06 * hampton2 has wanted good ref-counted objects for years. They would solve lots of problems with the GtkTreeModels.
11:29:24 <warlord> Well, we don't have that, yet.
11:29:35 <warlord> This is only the first step to that.
11:29:42 <hampton2> but its a big step in the right direction
11:29:47 <warlord> True.
11:29:52 <warlord> Okay, I'm gonna go merge..
11:30:58 *** Ardonik has quit IRC
11:31:31 *** Wilddev has joined #gnucash
11:31:31 *** gncbot sets mode: +o Wilddev
11:33:23 <warlord> hiya Wilddev
11:33:33 <Wilddev> hi warlord
11:35:56 <warlord> how are you?
11:36:49 <Wilddev> doing pretty good thanks
11:36:59 <warlord> Good!
11:37:05 <Wilddev> how bout u?
11:37:14 <warlord> I'm going to Dallas next weekend.
11:37:20 <Wilddev> woohoo
11:37:22 <warlord> :-D
11:37:25 <Wilddev> nice
11:37:31 <warlord> I'm excited.
11:37:53 <warlord> hampton: I merged locally.. Testing the build before I push it back up.
11:38:06 <Wilddev> I bet ur
11:38:15 <warlord> Wilddev: you would know ;)
11:38:44 <warlord> Oh, and work FINALLY ordered my new laptop. Unfortunately they delayed so long that the configuration I wanted wasn't available anymore.
11:39:34 <Wilddev> dell?
11:42:45 <warlord> thinkpad.
11:42:54 <Wilddev> aha
11:43:09 <Wilddev> I'm thinking about an Apple for my next one for work
11:43:30 <warlord> I originally spec'ed it out in January. It finally got ordered this week.
11:44:30 <warlord> hampton: "make check" worked in my merged trunk.
11:56:18 *** bonez39 has joined #gnucash
12:09:59 *** Ardonik has joined #gnucash
12:14:30 <warlord> okay.. committed.
12:15:07 *** bonez39 has quit IRC
12:21:55 <warlord> Hmm... I can get a 2x-Xeon Quad-Core 2.66GHz/ 1333MHz-FSB / 16GB / 2x1TB for about $8400.
12:22:18 *** BlackBsd has joined #GnuCash
12:27:07 *** Ardonik has quit IRC
12:29:02 *** Ardonik has joined #gnucash
12:37:12 <warlord> Phil's alive.
12:51:52 *** twunder has quit IRC
12:59:22 *** andi5 has joined #gnucash
12:59:22 *** gncbot sets mode: +o andi5
13:00:12 <andi5> fyi: r15847 compiles and starts up fine on windows... nice :) .... someone could add a channel topic though...
13:00:27 *** warlord changes topic to "Welcome! Don't meta-ask, ASK and WAIT. Gnucash 2.0.5 is released! The remove-group branch has landed - check it out! FAQ: <http://wiki.gnucash.org/wiki/FAQ>. This is a *publically-logged* channel, <http://lists.gnucash.org/logs/>"
13:00:37 <warlord> andi5: Good to hear! :)
13:00:46 <andi5> gotta go again :)
13:00:48 *** andi5 has quit IRC
13:01:07 <warlord> Now we can slowly work on getting all the objects migrated over to the GObject init(), dispose() and finalize() routines.
13:05:45 <Wilddev> what's the remove-group?
13:08:38 <warlord> Wilddev: removes the AccountGroup object and adds a ROOT Account
13:09:27 <Wilddev> which does?
13:10:32 <warlord> Um.. Nothing the user should see.
13:10:39 <Wilddev> oh ok
13:25:49 *** hampton2 changes topic to "Welcome! Don't meta-ask, ASK and WAIT. Gnucash 2.0.5 is released! FAQ: <http://wiki.gnucash.org/wiki/FAQ>. This is a *publically-logged* channel, <http://lists.gnucash.org/logs/>"
13:26:15 <hampton2> The remove-group commit is so last-month.
13:26:26 <warlord> We could mention the GObject commit ;)
13:58:50 *** twunder has joined #gnucash
13:59:57 <warlord> off to see Pan's Labrynth. TTYL>
14:00:00 *** warlord is now known as warlord-afk
14:42:42 *** pepe__ has joined #gnucash
14:50:51 *** pimienta has quit IRC
15:07:39 *** twunder has quit IRC
15:32:58 <Wilddev> hampton2: guess I'm not part of the 'in' crowd anymore
15:37:07 *** |gunni| has joined #gnucash
15:46:00 *** _gunni_ has quit IRC
15:48:20 *** Flimqy has joined #gnucash
15:49:22 <Flimqy> I had to delete and re-create a reconsiled transaction. The $ amount is the same. Is there any way to manually reconcile a single transaction now? Or alternately, is there any way to mass-unreconcile all the other transactions so I can reconsile them all normally again?
15:52:11 *** brandon has quit IRC
15:54:16 *** kielein has quit IRC
15:55:40 *** mcepl has joined #gnucash
16:08:10 *** sjc has joined #gnucash
16:17:02 <jsled> FYI, maybe for the future. http://blogs.gnome.org/view/ovitters/2006/11/12/0
16:17:19 <jsled> (auto-rejecting 'bad' stack traces from bug-buddy->bugzilla.)
16:23:55 <jsled> taxes suck.
16:24:42 *** Vantage has joined #gnucash
16:26:31 <Vantage> Hi, I'm trying to figure out how to handle currency conversion in gnucash. I can create an account for cash in wallet for the other currency, but when I take money from an account from one currency to the other, how do I handle the conversion?
16:41:20 <Vantage> nm, figured it out
16:44:53 *** Vantage has left #gnucash
16:47:42 <lasindi> jsled: Hi, I'm the student with the (one of the?) delimited file importer application. I was wondering if you were "satisfied" - if that's the right word - with my explanation for why I favored implementing our own widget. Would you rather use a tree view?
17:07:28 *** goibhni1 has joined #gnucash
17:14:20 *** goibhniu has quit IRC
17:15:06 <warlord-afk> Flimqy: No need. Next time you reconcile the account just ignore the starting balance.. Make sure the ending balance is correct, and just reconcile the old transaction along with all the new ones.
17:15:28 <Flimqy> warlord-afk: ah, good call. thanks.
17:15:32 *** warlord-afk is now known as warlord
17:19:15 <warlord> jsled: cool, that would be nice.
17:29:13 <hampton2> Should it be legal to call an xaccAccountGetXxx/xaccAccountSetXxx function and pass in a null account? The existing code silently returns. I'm thinking these should be g_return_val_if_fail checks.
17:31:01 *** mcepl has quit IRC
17:31:37 <warlord> Hmm...
17:32:04 <warlord> I think g_return_val_if_fail(GNC_IS_ACCOUNT(acct)) would be reasonable.
17:32:44 <hampton2> which I think should fail if acct is NULL
17:32:47 <hampton2> works for me.
17:32:50 <warlord> Right
17:33:01 <warlord> Also, I dont think your change to gncInvoice.c will actually work..
17:33:34 <hampton2> I thought it was semantically identical. It just broke one line into two.
17:34:17 <hampton2> gnucash wouldn't compile here without it.
17:34:35 <warlord> I'm not sure how that code worked before.. Unless that function just isn't used by anyone.
17:35:24 <hampton2> well, I'm not sure we should really be storing one object within another in the first place. Shouldn't it just store pointers to the other objects?
17:36:09 <warlord> Oh, those functions aren't even used.
17:36:26 <warlord> Well, now that we have a real class system we can move things around.
17:38:38 <warlord> i.e., we can make GncOwner into a real superset...
17:41:56 <jsled> lasindi: hey Ben. Your arguments make sense, yeah.
17:42:09 <jsled> lasindi: (ben? benny?)
17:42:38 <jsled> I'm just averse to custom widgets. Of course, it'd be even better if some widget could be shared/factored out of gnumeric's import or something.
17:43:47 <lasindi> Hi sorry, was afk for a little while there.
17:45:08 <lasindi> People usually call me Benny, but you can call me whatever you like. :-)
17:45:26 <hampton2> uhh, you don't want to say that......
17:45:35 <hampton2> ;-)
17:45:35 <jsled> lasindi: btw, so what do you see from the applicant side? Just your app and ... what else? anything about Gnucash or ...?
17:45:43 <jsled> mu ha ha ha ha.
17:46:04 <lasindi> Haha.
17:46:32 <lasindi> I see my app and your comment and cstim's (seems that's his name around here).
17:46:35 <lasindi> (and my comment)
17:46:52 <conrad> "Mind if we just call you Brian?" (possibly lost on everyone here - its a monty python quote, but what came to mind)
17:47:29 <jsled> conrad: heh.
17:48:02 <lasindi> I think I could pull the tree view out of Gnumeric at least at the beginning.
17:48:23 <lasindi> jsled: is there a way to draw vertical lines (instead of just spaces) between columns in a tree view?
17:48:47 <jsled> hampton: might know. Or chris, if he's around.
17:48:47 <lasindi> I researched and couldn't find anything, but I could be wrong ...
17:48:53 <hampton2> I don't recall any such option in GtkTreeView
17:50:05 <lasindi> Hm, ok. That's one of the things I kind of disliked about Gnumeric's wizard; it's relatively hard to see separation between columns.
17:50:41 *** nbinont has joined #gnucash
17:50:44 * hampton2 isn't sure what the "grid-line-pattern" style does on a GtkTreeView
17:51:34 *** goibhni1 has quit IRC
17:51:42 <hampton2> Whatever it is, its a style property which means you shouldn't set it, but let the system style (or a user's .gtkrc file) set it.
17:53:01 <lasindi> Yep.
17:53:24 <jsled> lasindi: it's funny, because I'm responsible for one of the custom widgets already in gnucash.,
17:53:39 <lasindi> Oh? Which one?
17:54:10 <hampton2> Bad jsled. Bad programmer. No doughnut.
17:54:22 <jsled> But the register is the one that always comes to mind, and it's a perrenial problem.
17:54:31 <jsled> lasindi: the Dense Calendar used by the SX code.
17:54:42 <jsled> hampton2: heh.
17:55:22 <jsled> well, I can somewhat justify the dense cal. And historically the register has been justifyable.
17:55:31 <jsled> justifiable, even.
17:56:10 <lasindi> Yeah, makes sense.
17:57:26 <lasindi> This might be a long shot, but there seems to be already a couple of GTK+ apps that need widgets for displaying spreadsheets; maybe if we implement one, it could get sent upstream and become a stock widget?
17:57:35 <lasindi> (kind of like the color selector)
17:59:29 <jsled> maybe not. When you say "spreadsheet", I think "editable array of cells", which is pretty much the GtkTreeView.
18:00:01 <jsled> The new stuff you're talking about doesn't exactly seem generally applicable.
18:00:21 <jsled> Though maybe it makes sense to do it as (optional) extensions of the gtktreeview itself.
18:00:26 <jsled> (rather than a new widget)
18:01:52 <lasindi> Ah, well, I'm talking more specifically about for displaying a delimited file (but not for editing),
18:02:42 <lasindi> but by "extensions" of gtktreeview, do you mean taking the gtktreeview source code and editing it?
18:04:00 <jsled> yeah, maybe.
18:04:37 <jsled> Maybe it just doens't work.
18:04:57 <jsled> I guess you don't want to drag columns between others, but to drag column headings on *top* of view columns.
18:05:02 <jsled> (to define their meaning)
18:05:05 <jsled> (or maybe their format)
18:05:13 <lasindi> Right
18:05:43 <jsled> Though, I still think you could fit this into a tree view, though it'd be a bit weird.
18:05:44 <lasindi> That could certainly be done with comboboxes, but I thought it might be more intuitive/quick to drag and drop.
18:06:04 <jsled> Yeah, It'd be like a two item root: "formatting" and "parsed-content".
18:06:04 <lasindi> Yeah
18:06:41 <jsled> The children of formatting would be the "column type/name" (combo) and data-type (combo; derived from the type)
18:07:00 <jsled> Maybe also an "ignored" (boolean/checkbox) child.
18:07:08 <jsled> Then all the parsed children would be the rows.
18:08:33 <jsled> Hmm. It might be interesting to define the idealized delimited-format-parsing Model, and start by adapting it into a Gtk{View,Model}. Then, focusing on the customized View secondarily.
18:08:53 <jsled> Though you're saying you already have the customized view from $other_project, so...
18:09:57 <lasindi> Yes, although it has lots of issues, not least of which is that it's written in Python at the moment.
18:10:04 <warlord> conrad: that's from Life of Brian, isn't it?
18:10:54 <jsled> warlord: No, I think it's the philosophers sketch.
18:11:11 <conrad> warlord: Meaning of life, I think (the Australian university scene)
18:11:17 <lasindi> I looked a lot through GnuCash code/mailing list archives to try and find bindings for other languages, but it appears that there was only some limited Perl support, which is now somewhat dead, right?
18:11:23 <conrad> They all kinda blur a bit.
18:12:14 <jsled> Jeez, I don't recall Meaning of Life at all. I only recall hearing that one on CD.
18:13:15 <jsled> lasindi: perl? Not so much. We extensively use bindings for guile/scheme ... historically through g-wrap, on trunk trhough swig.
18:13:17 <warlord> lasindi: there's the scheme bindings.
18:13:33 <jsled> There's also some not-in-tree python bindings for 2.0 ...
18:13:55 <warlord> but i invited that developer to donate those patches to us for inclusion.
18:14:02 <jsled> We never did hear back from that guy about using... yeah.
18:14:41 <lasindi> Did this guy just say, "I've got these bindings" and then disappear, or do you have actual source code?
18:15:21 <warlord> lasindi: I guess I still dont understand why we need a custom widget for CSV import. Why can't you just need to deal with the column -> name/type mapping and then feed it into the generic importer.
18:15:37 <warlord> er, "Why can't you just deal with the ..."
18:16:41 <warlord> (maybe you already discuss this somewhere?)
18:18:31 <lasindi> warlord: (It's discussed somewhat on my application's page, if you have access to it.) I think we certainly could get away with not making a custom widget; the only reason I'm kind of biased toward a custom widget is because the GTK implementations for this kind of functionality seem to be less intuitive than software with custom widgets.
18:19:30 <lasindi> e.g., comparing Gnumeric's GTK wizard with those in OpenOffice and Excel
18:20:33 * warlord has never played with the OOo or Excel CSV import.. (or the Gnumeric one, for that matter)
18:21:57 <warlord> lasindi: perhaps instead of showing the results for ALL lines in the file it only shows the first one?
18:22:10 <warlord> (I'm not sure if that makes life any easier for you)
18:22:57 <lasindi> That could be done; although once we can show one line, I think it's almost as easy to show them all.
18:23:42 <warlord> True.
18:24:23 <warlord> Another few things to keep in mind: It would be nice to have different CSV profiles, or ways to remember (multiple) previous CSV configurations.
18:24:37 <warlord> Don't forget to be able to handle stock transactions..
18:24:55 <warlord> And (a personal itch), it would be nice to be able to tell the importer than you're importing a split transaction.
18:25:17 <lasindi> Yes, good points.
18:26:14 <warlord> The mockup looks good, but I suspect you could do that with a GtkTreeView. C.f. the GncQueryList
18:27:13 <warlord> Oh, and obviously you need to choose the "source account", and eventually plug it into the generic import to choose destination accounts.
18:28:30 <lasindi> Right.
18:28:59 <warlord> I also dont understand why the user need to reorder the columns (except to match the column name/type to the actual data). One would think that if you have column headers, all you need to do is say: column X maps to data item Y of type Z
18:30:17 <lasindi> I wouldn't be reordering the columns, but I would be reordering the formats. Is that what you're asking?
18:30:59 *** Wilddev has quit IRC
18:31:41 <warlord> lasindi: okay, so it's not really "reorder" but more of a column mapping, mapping between the columns and transaction parameters/types.
18:31:49 <conrad> warlord, jsled: I was actually misquoting monty python - of course the quote should read "Mind if we call you Bruce?" and it was probably known as the philosopher's sketch (being the philosophy department of the University of Woolamaloo) - full marks to jsled for that one.
18:32:04 <lasindi> warlord: yup.
18:32:22 <jsled> Ah, Bruce! Of course.
18:32:28 <warlord> AHH! Bruce!
18:32:30 <warlord> Duh!
18:32:38 * jsled needs to find his Monty Python cds...
18:32:47 <jsled> Of all the things I've not ripped.
18:33:05 <warlord> lasindi: okay. Again, you might want to look at the GncQueryList.. We already have a display-style widget to display columnarized data.
18:33:33 *** nbinont has quit IRC
18:34:38 <lasindi> warlord: yes, I'm taking a look at that now; I think it actually might well be exactly what's needed.
18:35:28 <warlord> :)
18:36:40 <lasindi> By the way, how often does cstim come on to the channel? I've been idling on the channel for a couple weeks, but he doesn't appear to come in frequently. (I'm wondering since he would be my mentor if I get accepted.)
18:37:29 <warlord> He's been really busy with work, thesis, and family. He'll probably come back here more often in a couple more weeks.
18:37:36 <warlord> he DOES still read email
18:38:08 <warlord> anyways, I gotta run.
18:38:10 <lasindi> Ah, ok. :-)
18:38:10 *** warlord is now known as warlord-afk
18:38:11 <jsled> yeah, he took a specific, intentional and (hopefully) temporary leave from IRC.
18:38:15 <jsled> warlord: cheers.
18:38:26 <lasindi> See ya warlord.
18:38:52 <lasindi> Alright, that's cool.
18:41:08 *** rhowe has quit IRC
18:58:58 *** benoitg has joined #gnucash
18:58:58 *** gncbot sets mode: +o benoitg
18:59:08 *** benoitg has left #gnucash
19:09:57 *** magenpie has joined #gnucash
19:12:11 <magenpie> i'm having a problem creating new account...top level, bank account...has crashed repeatedly when i click OK
19:12:40 <magenpie> i can edit accounts ok
19:14:24 <magenpie> this is 2.0.5 in debian
19:18:13 <magenpie> i tried different names and it just 'worked' when i called it 'checking'
19:18:14 <magenpie> ugh
19:18:22 <magenpie> now maybe i can edit the name
19:18:55 <magenpie> yes
20:55:16 *** sjc has quit IRC
21:14:07 *** BlackBsd has quit IRC
22:08:05 *** magenpie has left #gnucash
22:58:33 *** Wilddev has joined #gnucash
22:58:33 *** gncbot sets mode: +o Wilddev
23:06:33 *** |gunni| has quit IRC