2007-08-18 GnuCash IRC logs

00:28:54 *** _lasindi has joined #gnucash
00:28:54 *** lasindi_ has quit IRC
00:37:29 *** dbr has quit IRC
00:48:35 *** lasindi_ has joined #gnucash
00:48:35 *** _lasindi has quit IRC
01:30:32 *** lasindi_ has quit IRC
01:31:15 *** lasindi has joined #gnucash
02:14:32 *** kielein has joined #gnucash
02:39:23 *** blahdeblah has joined #gnucash
02:41:39 <blahdeblah> Hi. My accountant tells me that for my company tax return he needs "a Profit & Loss return which requires a complete dissection of the deposits & withdrawals for the period".
02:41:39 <blahdeblah> What's the closest thing to this in the GnuCash reports?
03:29:17 *** Jaran|zZZZzz has quit IRC
04:45:13 *** Jaran has joined #gnucash
05:10:36 *** lasindi_ has joined #gnucash
05:16:49 *** lasindi has quit IRC
05:32:24 *** gunnicom_ has joined #gnucash
06:01:06 *** leafeater has joined #gnucash
06:02:38 <leafeater> hi, I have to work on Windows-2000 and am trying to use my SEB Chipcard. The qt3-wizard just disappears when i try to read a user from the chipcard. Can anyone help?
06:04:33 *** sorush20 has left #gnucash
06:10:10 *** lamp02 has joined #gnucash
06:46:05 *** lamp02 has left #gnucash
07:04:10 *** cortana has quit IRC
07:07:11 *** cortana has joined #gnucash
07:35:39 *** andi5 has joined #gnucash
07:35:39 *** gncbot sets mode: +o andi5
07:50:13 <leafeater> hi, I have to work on Windows-2000 and am trying to use my SEB Chipcard. The qt3-wizard just disappears when i try to read a user from the chipcard. Can anyone help?
07:53:23 <andi5> leafeater: you mean online banking? quoting http://wiki.gnucash.org/wiki/AqBanking#HBCI_Security_Media: "Chipcards are not supported on Windows." :-(
07:55:42 *** gunnicom| has joined #gnucash
08:04:35 *** gunnicom_ has quit IRC
08:05:28 <leafeater> oh, thanks andi5, i believed the statement, that the windows port was finished. Well, anyway, where will be announced, when the chipcard support is finished for windows?
08:07:04 <leafeater> I might even like to try to help, if there is not already someone working on it...
08:09:03 <andi5> leafeater: well, that is aqbanking/libchipcard and all we do is that we compile against, use and ship it :-) ... that said, you will need to follow development on aqbanking-devel or libchipcard-devel (http://www.aquamaniac.de/aqbanking/ , http://www.aquamaniac.de/libchipcard/)
08:11:01 <andi5> leafeater: you might start with reading the whole wiki page i linked to above.. maybe it helps to track down the problem a bit
08:19:18 <leafeater> kk
08:25:28 <leafeater> how can i follow (and what is) aqbanking-devel and libchipcard-devel? They sound like linux package names, but how can i follow development on them (website, forum, email-list, ...)?
08:25:56 <andi5> those are mailing lists
09:55:05 <leafeater> k i have to go now. I subscribed to the libchipcard list and I will look into it.
09:59:09 *** andi5 has quit IRC
10:01:27 *** leafeater has left #gnucash
10:30:43 *** twunder has joined #gnucash
10:39:18 *** twunder has quit IRC
10:41:14 *** blahdeblah has quit IRC
11:52:21 *** Def has joined #gnucash
13:09:38 *** andi5 has joined #gnucash
13:09:39 *** gncbot sets mode: +o andi5
13:41:39 *** gunnicom| has quit IRC
13:48:30 *** snvv has joined #gnucash
14:02:32 *** snvv has left #gnucash
14:06:51 *** andi5 has quit IRC
14:07:48 *** andi6 has joined #gnucash
14:07:48 *** gncbot sets mode: +o andi6
14:51:47 *** andi6 has quit IRC
14:53:05 *** lasindi_ has quit IRC
14:55:23 *** lasindi_ has joined #gnucash
16:03:28 *** cyclope has joined #gnucash
16:03:35 <cyclope> hi
16:03:42 <cyclope> could anyone help me?
16:05:44 <cyclope> what should i do to run gnucash on Slackware 11.0 without GNOME? on any older version i successfully run it without GNOME
16:05:54 <cyclope> or what should i install to run it
16:05:57 <cyclope> thanks in advance
16:06:25 <jsled> cyclope: http://wiki.gnucash.org/wiki/Slackware#Slack_11_.2B_GnuCash_2
16:06:49 <jsled> You don't need to use GNOME as your desktop environment, but we do depend on GNOME libraries.
16:06:54 <jsled> So they have to be installed.
16:08:33 <jsled> http://svn.gnucash.org/repo/gnucash/trunk/packaging/gnucash-2.1.x.ebuild has a good list of the packages we require; though they're gentoo package-name/version tokens, they're just the normal package name/versions.
16:10:16 <jsled> And we've depended on gnome for many releases, now ... since 1.6, at least, I think even back to 1.4 or earlier.
16:10:34 <jsled> In any case, sorry to run, but I'm just passing through...
16:11:25 <cyclope> ok, thanks. i will try older version or dropline for the latest one
16:11:52 *** cyclope has quit IRC
16:58:13 *** kielein has quit IRC
17:44:49 *** Jaran has quit IRC
18:11:08 *** twunder has joined #gnucash
18:14:39 *** twunder has quit IRC
19:30:08 *** pdc has quit IRC
19:30:27 *** pdc has joined #gnucash
20:27:28 *** warlord-afk is now known as warlord
20:27:43 <warlord> @tell blahdeblah The Income Statement
20:27:43 <gncbot> warlord: The operation succeeded.
21:05:02 *** warlord is now known as warlord-afk
21:26:32 *** andi5 has joined #gnucash
21:26:33 *** gncbot sets mode: +o andi5
22:04:52 *** twunder has joined #gnucash
22:15:11 <chris> go andi5! :)
22:15:18 <andi5> hehe
22:21:08 <andi5> chris: can do safe me 5 minutes and tell me why i cannot see the rate, type and notes columns? :)
22:22:18 <andi5> oh, btw... what is the preferred way to enter a new transaction? enter will probably only start editing the selected cell, right? .. tabbing out?
22:29:02 <chris> andi5: because they are not in col_list.
22:29:37 <andi5> oh, right
22:30:03 <chris> The view is supposed enumerate and handle all possible columns, but individual instantiations of the view may choose to use whatever list of columns they want.
22:30:52 <chris> related to that, though... I was never really sold on the double line mode.
22:31:33 <chris> I don't know if you can stack cellrenderers. I think you couldn't when I first looked, but that was a long time ago.
22:32:17 <chris> anyway, I'm just not convinced that double-line mode is worth it, considering that the user can re-order, hide/show, and resize all columns as desired.
22:32:50 <chris> About entering... ideally enter should enter the transaction.
22:32:55 <andi5> ah... i think i remember what that mode is again
22:33:27 <chris> yeah, I think it's very rarely used.
22:33:34 <andi5> do you have any idea how many users make use of it?
22:33:43 <andi5> well..
22:34:00 <andi5> do not even mention that it was removed ;-)
22:35:13 <chris> however, there are some keynav issues remaining. I *think* that the strategy of automatically beginning the edit on keynav is going to end poorly.
22:35:32 <andi5> btw, feel free to edit the announcement of 2.2.1... honestly, i have not yet looked at it at all and i will probably need your help tomorrow :)
22:35:51 <chris> Instead, I think it's more promising to try to catch the first edit keystroke and begin edit then.
22:36:34 <andi5> why not catch key presses that fall through all other handlers?
22:37:14 <chris> I don't know the details, that might work.
22:37:36 <andi5> so enter is "commit txn" again then?
22:38:16 <andi5> oh, another one... is it intended that clicking a txn from a list of txns with the same date moves it to the end of that list?
22:38:26 <andi5> that is pretty... confusing
22:47:31 *** twunder has quit IRC
22:48:09 <chris> re: enter, yes, I think that makes sense, and is also current behavior.
22:49:32 <chris> andi5: re sorting ... No, that's just horrible, and I know exactly what causes it.
22:49:55 <andi5> then go fix it ;-)
22:50:34 <chris> The GtkTreeModelSort sorts by the key you specify...
22:50:56 <chris> But among items with equal keys, the sort order is undefined.
22:51:41 <chris> however, the actual implementation is that the order depends on which rows last emitted events.
22:52:21 <chris> So, the process of moving the blank split around triggers events, which reorders transactions with equal sort keys.
22:52:23 <andi5> btw... the query did not sort for us?
22:53:25 <andi5> ah, i see... we have clickable headers :)
22:53:34 <andi5> but... well...
22:53:38 <andi5> (getting late here)
22:53:53 <chris> I don't know, it probably did, but enabling the user to sort as desired seemed pretty easy.
22:54:09 <chris> yeah, here too. :)
22:54:29 <andi5> i just want to see red balances...
22:54:46 <chris> oh, that's *easy*. :)
22:54:56 <chris> exercise left to reader.
22:55:08 <andi5> pah
22:55:23 * andi5 did not read that
22:55:53 <chris> btw, does it crash much?
22:56:07 <andi5> crash?
22:56:27 <andi5> i have not yet seen any since i started compiling again
22:56:47 <chris> yeah, that's one of the design goals - not crashing. ;)
22:56:53 <andi5> cool
22:57:26 <chris> showing transactions is in the top 5, too.
22:57:43 <andi5> boiling water?
22:58:08 <chris> no, just fetching it.
22:58:11 <andi5> ah, i see
22:58:33 <andi5> chris... it does not compile.... that is unfair
22:58:59 <chris> what?
22:59:17 <andi5> i made a change, but it does not compile anymore....
22:59:27 <andi5> there should be some law against that...
22:59:35 <andi5> murphy?
22:59:52 <chris> I think that's the law *for* that.
22:59:57 <andi5> arrg
23:02:22 <andi5> ... just do not add the column twice
23:05:16 <chris> andi5: when do you think we'll branch for 2.2.x?
23:05:25 <andi5> i have no clue
23:05:34 <andi5> do we need the branch now?
23:06:10 <chris> well, it doesn't seem like there's lots of trunk activity.
23:06:24 <chris> but I don't know if it matters.
23:06:27 <andi5> chris
23:06:33 <chris> andi5
23:06:57 <andi5> what about branching at some point and then making a big step like merging rr into trunk?
23:07:11 <andi5> just to avoid what seems to have happened with the gnome2 transition?
23:09:00 <chris> well, I think that in conjuction with either a configure-time flag, or a preference, or some other conditional, that's reasonable.
23:09:03 <andi5> maybe one should not compare these two situations... i joined after the merge of the gnome2 branch, so who am i :)
23:09:51 <chris> I really didn't know if a gtktreeview-based register was feasible when I started the branch.
23:10:08 <chris> but, now I think it's pretty obvious that it it.
23:10:11 <chris> *is.
23:10:46 <andi5> another reason to do this is that the register is the *heart* of gnucash... people will have to try it and make suggestions/bug reports and whatever.... yeah, and maybe help us to improve/complete it
23:11:08 <andi5> i think putting rr onto trunk makes it just much more visible
23:12:31 <chris> I agree. I think it's a good move for the trunk after the branch for 2.2
23:13:13 <chris> btw, if you want to side-by-side compare, you can still open regular registers from the budget's CoA.
23:13:14 <andi5> great...
23:13:23 <andi5> i know, i know :)
23:13:36 <chris> just making sure. :)
23:13:55 <andi5> it is both, frustrating and encouraging, from my point of view
23:14:37 <chris> for me, it's encouraging to see it so far along, and frustrating to have no time for it.
23:18:22 <andi5> two commits more for you to review
23:20:07 <chris> if they're not quick, I won't get to them tomorrow after noon, at the earliest.
23:20:36 <andi5> well, i guess r-r is not that time-critical :)
23:25:32 <chris> did you mean to drop the g_return_val_if_fail(g_list_length(cols) > 0, FALSE)?
23:25:45 <andi5> yes
23:26:11 <andi5> g_list_find will not find anything, and traversing the list once is enough, i guess :)
23:26:24 <andi5> (ok, it is not extremly big)
23:26:35 <andi5> or long
23:27:19 <chris> ok.
23:27:40 <chris> you know, I'm actually not a bug fan of the way the numeric columns are styled.
23:27:45 <chris> *big
23:27:59 <andi5> is there any reason not to print the currency symbol in the balance column?
23:28:33 <chris> I know that the gtk examples show you (and the api allows you) how to set the color and visibility using a model column.
23:29:36 <chris> but, I think it makes more sense to use the model only for the data, and to style the column in the cdf.
23:29:56 <chris> It's really no harder, *if* you're using a cdf anyway.
23:30:19 <chris> I think the examples don't do it that way because using a cdf is considered advanced.
23:31:31 <andi5> how can a cdf control visibility?
23:32:04 <chris> I notice that you didn't actually add a column to model yet.
23:32:16 <andi5> yep
23:32:54 <chris> Also, realize that if you do it the way you've started, then two views can't use different styles if they share the model.
23:33:22 <chris> hm, visibility?
23:35:01 <andi5> yes... and i think alignment and color should be the same among all views...
23:35:24 <andi5> i am not so sure about the selection of visible columns, though
23:35:47 <andi5> but that is defined somewhere else, right?
23:36:11 <chris> gconf, I believe.
23:36:45 <andi5> btw... you chose a different solution than gnc-tree-view-account by creating a view with a subset of the columns, instead of adding all columns and then set visility properties... is that right?
23:37:03 <andi5> yes, gconf... and in under different prefixes
23:37:41 <andi5> or maybe i did not read carefully enough
23:37:49 <chris> well, I think it does both...
23:38:35 <chris> The idea being that whoever creates the gtvt can decide what columns it may have.
23:38:35 <andi5> so what do you think the cdf should take care of?
23:38:54 <andi5> yeah... i think i got that now
23:38:58 <chris> but the user can still hide/show columns.
23:39:31 <chris> I think the cdf should handle color.
23:40:20 <andi5> ok... i will see whether that works for me
23:40:26 <chris> I don't see what a model column has to do with column visibilty, though.
23:40:37 <chris> but, I *am* tired.
23:41:38 <chris> that thing controls the visibility of *rows*, doesn't it?
23:41:48 <andi5> maybe not column visibility, but cell visibility... but maybe that can be handled by the cdf as well
23:42:48 <andi5> http://svn.gnucash.org/trac/browser/gnucash/branches/register-rewrite/src/gnome-utils/gnc-tree-view.c#L1905
23:43:47 <andi5> are gconf queries cached somehow?
23:43:49 <chris> what does that do?
23:44:22 <andi5> good question... i will not answer it today ;-)
23:46:46 <chris> basically, I think the gnc_tree_view_add_{custom}_column basically convenient if you don't want/have to use a cdf.
23:47:22 <chris> But that if you are going to use a cdf, you already have an easier way to have more control over style.
23:48:04 <chris> ok, I'm barely coherent, so I should sleep now.
23:48:26 <andi5> so... an easy Q... the user has a preference to decide whether negative values should be red, right?... is it reasonable to query gconf (maybe not the server) several times each row?
23:49:12 <chris> maybe.
23:50:53 <chris> I would try it and see.
23:54:02 <chris> you really don't want to cache it in the model, so you're getting it at least once per model row anyway.
23:54:59 <chris> and there's no signficant difference between once per model row, and twice per view row.
23:56:12 <andi5> is there any reason why you chose V for void instead of v?
23:56:51 <chris> In fact, it may be even better to have it in the view, because the model may eagerly query it, while the view might only query for the visible cells.
23:57:33 <andi5> hm...
23:57:48 <chris> 'V' where?
23:58:31 <andi5> http://pastebin.ca/raw/662899
23:59:33 <chris> no idea. looks good.