2007-12-13 GnuCash IRC logs

00:38:14 *** thechris has joined #gnucash
00:38:50 <thechris> is there a way to get quotes on mutual funds in gnucash?
00:48:55 <warlord> if there's a supported quote source.
00:49:27 <thechris> i can type the symbol into yahoo, google and they work there
00:49:46 <thechris> but yahoo doesn't list where its from
00:49:54 <thechris> just "VFSTX"
00:50:31 <thechris> and FUND:VFSTX (or NYSE, AMEX, NASDAQ, or MUTF doen't work either)
00:50:55 <thechris> and no namespace is invalid
01:09:51 *** Renato has joined #gnucash
01:09:57 <Renato> Hi..
01:10:52 <Renato> I want to start tweaking with gnucash.. I've already downloaded source and wanted to know what IDE environment you guys recommend to jump start...
01:11:58 <warlord> thechris: yahoo VFSTX works for me. Did you configure that in the security editor?
01:12:15 <thechris> warlord: yes, what did you set it up as?
01:12:17 <warlord> Renato: use whatever you want. most of the dev's just use emacs.
01:12:33 <warlord> thechris: quote source: yahoo. symbol VFSTX
01:12:44 <thechris> what namespace?
01:12:52 <warlord> Doesn't matter
01:13:04 <warlord> Namespaces are for your sake, not for gnucash.
01:13:10 <warlord> But I'd use "FUND"
01:13:17 <thechris> is it case sensitive?
01:13:27 <Renato> warlord: ok... so I suppose that's a green light for my favorite ide: VI :-)
01:13:40 <warlord> thechris: yes
01:13:43 <Renato> (you see... been too long developing java...)
01:13:50 <warlord> er, the symbol is.
01:13:55 <warlord> the namespace doesn't matter
01:14:02 <warlord> Renato: sure.
01:14:33 <thechris> hmm, interesting. well that seems to work now, thanks
01:15:05 *** HelgeS has left #gnucash
01:15:08 <warlord> you're welcome
01:15:58 <Renato> one last question: once I've done packaging some fixes/improvements/whatever.. What's the best way to propose a merge on the code svn ? Should I post it to a mailing list and ask one of the maintainers to look at it ?
01:17:32 <thechris> is there a good way to track profit/loss from securities in gnucash?
01:18:43 <thechris> eg, if my current value exceeds or doesn't exceed my total purchased value, does this show up anywhere?
01:21:26 *** Renato has left #gnucash
01:23:05 <warlord> thechris: umm.. Advanced Portfolio Report?
01:23:43 <warlord> @tell Renato send patches to the -devel list, or file a bug report in bugzilla and attach the patch there? Small patches are better than large ones.
01:23:43 <gncbot> warlord: The operation succeeded.
01:27:13 <thechris> thanks
01:27:17 *** thechris has quit IRC
01:50:21 *** Zoolooc_ has quit IRC
01:52:32 *** warlord is now known as warlord-afk
03:11:57 *** ErKa has joined #gnucash
04:57:27 *** tim_abell has joined #gnucash
05:19:05 *** cortilap has joined #gnucash
06:37:35 *** andi5 has joined #gnucash
07:02:10 *** kling0n has joined #gnucash
07:27:33 *** warlord-afk has quit IRC
07:46:03 *** IanL has joined #gnucash
09:16:13 *** twunder has joined #gnucash
09:17:05 *** tjanssen has joined #gnucash
09:28:19 *** kielein has joined #gnucash
10:05:15 *** twunder has quit IRC
10:05:28 *** twunder has joined #gnucash
10:19:38 *** warlord has joined #gnucash
10:19:38 *** gncbot sets mode: +o warlord
10:27:26 *** IanL has quit IRC
10:35:58 *** donfede has joined #gnucash
11:10:13 *** ErKa has quit IRC
11:39:43 *** warlord has quit IRC
11:54:38 *** dbreiser has joined #gnucash
12:34:58 *** andi5 has left #gnucash
12:37:32 *** earthwalker has joined #gnucash
12:38:08 <earthwalker> Have a question about moving money between two different bank accounts.
12:42:34 <jsled> I'll guess it's about timing?
12:43:43 *** warlord has joined #gnucash
12:43:44 *** gncbot sets mode: +o warlord
13:02:56 <earthwalker> I have my personal bank account and we have a household bank account. They are on different gnucash accounts. So when I move money from my personal account to the house account how do I track it?
13:03:03 *** cortilap has quit IRC
13:04:06 <jsled> It'd be Assets -> Expense on the "from" set of books, and Income -> Assets on the "to" set.
13:12:32 *** tim_abell has quit IRC
13:18:28 *** ErKa has joined #gnucash
14:19:35 *** andrewsw-afk is now known as andrewsw
14:42:14 <earthwalker> Thank you very much.
14:49:41 *** warlord is now known as warlord-afk
15:00:52 *** earthwalker has left #gnucash
15:00:55 *** kielein has quit IRC
15:06:35 *** camgirl29 has joined #gnucash
15:10:49 *** chris has quit IRC
15:41:07 <andrewsw> I'm digging into this "Search dialog vanishes" problem. I don't know jack about gui code, but
15:41:26 <andrewsw> the folks at Xmonad tracked part of it down and it looks like it's absurdly simple.
15:42:32 <andrewsw> there is a gtk_widget_hide (sw->dialog); right after the gtk_widget_show_all (sw->dialog);
15:42:41 <jsled> bwahahha...
15:42:44 <andrewsw> I"ve commented out the _hide and now it just works...
15:43:07 <andrewsw> heh
15:43:41 <andrewsw> it's in src/gnome-search/dialog-search.c about line 858.
15:43:59 <andrewsw> the dev at xmonad says "there must be someone reason gnucash did that..."
15:44:44 <jsled> svn blame...
15:44:58 <jsled> Maybe the commit logs/history describe why.
15:46:56 <andrewsw> ignoring the possibility that there is a valid rason to do this _hide, am I dreaming that it's so simple?
15:47:07 <jsled> nope.
15:49:08 *** warlord-afk is now known as warlord
15:51:12 <warlord> Assuming it's the one that has "Hide the 'new' button..." comment right after it.. looks like andi5 added it on 5/13
15:52:07 <warlord> (this was r16078)
15:52:12 <andrewsw> yes, that's the one.
15:52:18 <andrewsw> line 853
15:52:34 <warlord> svk blame didn't give me line numbers.. But it's some of the only code *I* didn't write. ;)
15:52:44 <andrewsw> heh heh
15:54:03 <andrewsw> ugh. windows problem.
15:55:29 <andrewsw> basically what xmonad (and others) appear to highlight is that it maps and unmaps itself a couple times. This highlights a bug in these WM's, but they also call the behavior "odd".
15:57:00 <elb> gnucash has a whole raftload of odd window behaviors, as best I can tell :-P
15:57:05 <andrewsw> wouldn't it be better to move the gtk_widget_show_all to the end instead of showing it, hiding it, and then reshowing it?
15:57:12 <elb> even when I tell it not to remember window size and position, it moves itself around
15:57:43 <andrewsw> yeah. use a tiling wm and that problem goes away ;), just can't search.
15:58:26 <elb> heh
15:58:35 <elb> I guess you just have to decide what's more important ;-)
15:58:46 <andrewsw> that's what links in reports are for...
15:59:32 <warlord> andrewsw: well, there are some widgets that want to be hiddin the "final" showing.. Which is why it does: gtk_widget_show_all(dialog), gtk_widget_hide(dialog), ..... , gtk_widget_show(dialog)
16:00:57 <andrewsw> show_all turns on everything, right? but then we don't want to actually see it yet, so hide it?, tweak stuff, then show it for real?
16:01:41 <warlord> Right.
16:01:50 <warlord> And that's a PERFECTLY reasonable way to implement it.
16:01:58 <warlord> If the WM cannot handle that, the WM is broken.
16:02:12 <andrewsw> sure. I get it.
16:02:49 <elb> it's not very X11-friendly
16:03:06 <elb> (because you map the dialog, then unmap it and remap only parts -- lots of RTTs)
16:03:45 <elb> (not that any GNOMEish application is friendly to X11, really, and most people can't tell because they never use nonlocal connections)
16:04:46 <andrewsw> is it a matter of show_all, hide dialog, hide one element, show dialog as opposed to
16:04:49 <warlord> Well, there's no other way to have a base dialog, make everything in the base "showable", and then add more showable and hidden widgets before actually showing the whole dialog.
16:05:02 <andrewsw> nm.
16:05:09 <andrewsw> okay. interesting.
16:05:27 <warlord> I.e, there is now "gtk_widget_show_all_children()"
16:05:31 <warlord> s/now/no
16:05:54 <andrewsw> aha. just what I was thinking.
16:06:44 <andrewsw> with that you could show_all_children, _hide one or two children, and then _show(dialog).
16:07:13 <elb> you can iterate children, though, right, which means that writing such a function would not be too difficult
16:07:13 <warlord> Right
16:07:25 <elb> GtkContainer has a child iterator function
16:07:58 <warlord> elb: sounds like a great patch to send to us: void gnc_gtk_widget_show_all_children(GtkWidget*)
16:08:20 <elb> heh if I were sending a patch to improve remote X11, I'd be starting wit hthe ledger ;-)
16:08:29 <jsled> do both! :)
16:08:38 <elb> (it apparently has a lot of races -- while it glitches occasionally locally, it glitches a *lot* over remote X11)
16:09:05 <elb> sometimes I think I'm the Last Real X11 User ;-)
16:09:59 <warlord> Hehe
16:10:25 <warlord> seriously, if you want to get your fingers wet but don't have a lot of time, you could send in this API implementation and then slowly add patches to various dialogs to implement it this way.
16:11:13 <elb> I'm looking at it now, it wouldn't be too tough
16:11:28 <elb> ideally yo uwant to start at the innermost widgets first, becuas ethat minimizes repacking
16:11:35 <elb> so some bookkeeping is required
16:11:41 <elb> but ... there is indeed a gtk_container_foreach
16:13:45 <warlord> The way I'd do it is actually make it a "hide_parent_and_show_all_children" API.. So you can start with "gtk_widget_hide(widget);" and then just gtk_container_foreach(widget->children, gtk_widget_show_all)
16:13:52 <warlord> ... or something like that.
16:13:57 <elb> yeah
16:14:04 <elb> I wonder how smart Gtk+ is about that
16:14:13 <elb> (the Gtk+ guys apparently NEVER use remote X11)
16:14:45 <elb> gnucash is actually one of the few Gtk+ applications I use remotely
16:14:56 <elb> which is probably a compliment to gnucash ;-)
16:15:08 <elb> it's So Good that I put up with crapola remote performance to use it anyway ;-)
16:15:27 <andrewsw> I use it remotely now and then with no real problems (local lan though).
16:15:36 <elb> local lan is sluggish, but not terrible
16:15:50 <elb> I regularly use it from consumer residential connection to consumer residential connection
16:15:54 <elb> which is ... not fast
16:15:59 <andrewsw> eww...
16:16:59 <warlord> To use it remotely I use VNC. ;)
16:18:00 <elb> boo
16:18:07 <warlord> :-P
16:18:08 <elb> compromises like that are what make remote X11 suck!
16:18:11 <elb> ;-)
16:18:30 <warlord> I have many other reasons to use VNC other than that..
16:18:35 <elb> I used to use vnc a lot, but now I almost never do ... virtually all of the remote apps I use are either console or fast X11
16:18:52 <warlord> That machine also runs VMWare Workstation and I need access to the VM Console. .
16:19:09 <andrewsw> I just keep my whole world in screen and ssh all over the place.
16:20:18 <warlord> screen doesn't help you with X apps.
16:20:53 <andrewsw> export DISPLAY=...
16:21:09 <andrewsw> seriously though, the X apps I use are firefox and gnucash.
16:21:57 <warlord> Have you ever tried firefox over remote-X?
16:22:02 <andrewsw> heh heh
16:22:03 <andrewsw> LOL
16:22:08 <warlord> Anyways, I've got a talk to give in 10 minutes... BIAB.
16:22:13 <andrewsw> have fun!
16:23:08 <warlord> Will do.
16:23:11 *** warlord is now known as warlord-afk
16:28:04 <andrewsw> well, despite all that, removing that gtk_widget_hide seems to work just fine. I'm not really sure what the issue is...
16:29:10 *** kling0n has quit IRC
16:30:49 *** dbreiser has quit IRC
16:56:59 *** ahz has joined #gnucash
16:57:28 <ahz> I run "gnucash --debug" from terminal and have (gnc:debug "hello") in a .scm report, but terminal shows nothing
16:57:52 <jsled> yup!
16:58:09 <andrewsw> ahz: you have to set a log level for debug.
16:58:12 <jsled> try {gnucash --log gnc.scm=debug}
16:58:32 <jsled> (also, maybe use {--logto stderr} rather than consulting /tmp/gnucash.trace, but that's preference)
16:58:36 <andrewsw> and --logto stdout
16:59:03 <jsled> "--debug" actually sets the "base" loggers to info-level, since debug-level for gnucash is just *way* too much.
16:59:19 <jsled> Really, gnc:debug should probably map to "info"...
17:03:43 <andrewsw> later
17:03:47 *** andrewsw is now known as andrewsw-afk
17:06:17 <ahz> Hmm.. With {gnucash --log gnc.scm=debug --logto stdout} I do not see gnc:debug or gnc:warn in budget.scm. I also tried gnc.scm=info
17:06:55 <ahz> nevermind
17:06:59 <ahz> I just don't know my Scheme
17:07:04 <jsled> ah. heh.
17:07:45 <jsled> If you get tired of the command-line options, you can create an equivalent ~/.gnucash/log.conf file.
17:08:04 <ahz> Perhaps someone could update http://wiki.gnucash.org/wiki/Custom_Reports because it recommends {gnucash --debug} which does not work--even if you know Scheme :)
17:08:14 <ahz> thanks for the tips
17:08:25 <jsled> you're welcome.
17:24:29 <ahz> I try to subtract $70 from $500 in a report using gnc-numeric-sub-fixed, but it fails. I think it's because one has denom 100 and the other denom 1.
17:28:52 <jsled> why not just gnc_numeric_sub(...)?
17:28:55 *** jenny20 has joined #gnucash
17:30:04 <jsled> With, say, GNC_HOW_DENOM_REDUCE
17:30:59 *** jenny20 has quit IRC
17:31:52 <jsled> ahz: gnc-numeric.h suggests [[[
17:31:53 <jsled> To use traditional rational-number operational semantics (all results
17:31:53 <jsled> are exact and are reduced to relatively-prime fractions) pass the
17:31:53 <jsled> argument GNC_DENOM_AUTO as 'denom' and
17:31:53 <jsled> GNC_HOW_DENOM_REDUCE| GNC_HOW_RND_NEVER as 'how'.
17:31:53 <jsled> ]]]
17:32:16 <ahz> hmm, let me see
17:32:32 <jsled> (hmm. why isn't gnc-numeric.h in the doxygen output?)
17:32:34 <ahz> Is "GNC_HOW_DENOM_REDUCE| GNC_HOW_RND_NEVER" valid in scheme?
17:33:13 <jsled> Well, it's a C-idiom statement, so not directly.
17:33:47 <jsled> I don't know what the equivalent bitwise operator in scheme is… (bitwise-or …) maybe?
17:35:42 <ahz> bitwise-ior it seems
17:37:07 <jsled> Pfft. I was looking in the wrong place. http://cvs.gnucash.org/docs/HEAD/group__Numeric.html
17:43:54 *** twunder has quit IRC
17:52:10 <ahz> Thanks. I am closer, but when operand denoms are both 100 the result is something like $37 + 7 / 25
17:52:59 *** warlord-afk is now known as warlord
17:53:15 <warlord> andrewsw-afk: perhaps there were other changed in that changeset?
17:56:44 <ahz> Are there ways to display splash and to disable business-gnome without recompiling?
17:58:39 <ahz> jsled: When I used LCD instead of REDUCE, it looks much better. Thanks, I will double check and then have a patch for http://bugzilla.gnome.org/show_bug.cgi?id=347274
18:07:02 <ahz> Will my patch for http://bugzilla.gnome.org/show_bug.cgi?id=347274 automatically be reviewed, or do I need to do something?
18:07:41 <warlord> ahz: you can disable business-gnome by rm'ing the module.
18:08:06 <warlord> ahz: an email reminder to -devel would help us remember to look at your patch.
18:10:02 <ahz> warlord: thanks, I used chmod a= and now GC loads 3.4sec faster :)
18:12:05 <warlord> okay.
18:22:26 <andrewsw-afk> ahz: in scheme, just remove the "HOW" for rounding. GNC_HOW_RND_NEVER becomes GNC-RND-NEVER
18:22:45 * andrewsw-afk is drive by commenting
18:23:40 *** Zoolooc has joined #gnucash
18:50:07 <ahz> andrewsw-afk: thx, looks much cleaner
18:52:11 *** chris has joined #gnucash
18:52:11 *** gncbot sets mode: +o chris
19:14:29 <ttick> has anyone written a PDF to QIF translator? Just thought I'd ask before I write one.
19:14:55 <jsled> PDF? No. There's no clear mapping of PDF to QIF.
19:15:21 <jsled> I'd probably try to use something like pdf2txt to get something manipulable, then map it over to QIF.
19:15:35 <ttick> well, it would have to be tailored to different banks/cc companies . . . but certainly possible.
19:15:53 <ttick> right, that's what I was planning: pdf2xml -> perl -> QIF
19:16:09 <ttick> or Java, or XSLT is probably a better choice.
19:16:22 <jsled> Yeah ... it'd be more "Bank of America's late 2007 formatted checking account PDF"2qif. :)
19:16:28 <ttick> just thought I'd ask before I spent all that time writing it.
19:16:29 <jsled> XSLT is never a better choice. :p
19:16:34 <jsled> Go for it.
19:17:02 <jsled> It's hard to justify Java as a good choice for this, too.
19:17:04 <ttick> I think I'll make it my January project. Chase CC -> QFX or something.
19:17:11 <jsled> :)
19:17:37 <ttick> Well, it depends what the resulting XML looks like. But XSLT is the logical choice -- though maybe not the quickest to write.
19:17:45 <jsled> Note that there's a CSV importer in svn ... so getting to either QIF or CSV is good.
19:17:56 <ttick> ah, right.
19:18:13 <ttick> good point. Though one is probably as good as the other.
19:18:39 <ttick> any benefit to going for QFX instead of CSV? (more complex transactions available?)
19:18:51 <jsled> At least the semantics of QIF and *FX are related to accounting.
19:19:15 <jsled> I don't know much about the OFX/QFX format. I believe they can represent more complex transactions, true.
19:19:32 <ttick> unfortunately, Chase won't give me a QFX for anything earlier than 90 days. :( Their support email: "You'll have to manually type it in."
19:19:38 <jsled> nice.
19:19:50 <ttick> OFX (Open Source QFX?)
19:20:15 <jsled> no no ... the other way around. OFX = Open Financial Exchange. QFX = Quicken-bastardized Financial Exchange. :)
19:20:27 *** donfede has quit IRC
19:20:33 <ttick> ah.
19:20:55 <jsled> (though "Open" in this case is the same as any other "Open" format micros~1 is affiliated with ... OFX was spec'd by MS, Quicken and CheckFree.)
19:21:38 <ttick> gotcha.
19:23:17 <jsled> Anyways; gotta run.
19:23:21 <ttick> latter
19:23:24 <ttick> later
19:31:59 *** ahz has quit IRC
19:33:58 *** ErKa has quit IRC
19:44:09 *** joerlend__ has quit IRC
21:29:42 *** twunder has joined #gnucash
21:37:28 *** twunder has quit IRC
21:45:02 *** warlord has quit IRC
22:15:26 *** Zoolooc_ has joined #gnucash
22:20:11 *** joerlend has joined #GnuCash
22:24:45 *** Zoolooc has quit IRC
22:40:05 *** tjanssen has quit IRC
22:43:04 *** nbinont has joined #gnucash
22:50:44 *** pac1 has joined #gnucash
22:51:47 <pac1> tim_abell, I'm at the same place.
22:52:31 <pac1> Is 2.2.1 trunk?
22:56:40 *** zinmirai has joined #gnucash
23:01:31 *** zinmirai has quit IRC
23:21:38 <jsled> pac1: trunk is close, but not quite.
23:43:45 <ttick> okay, looks like the XML of the PDFs are going to be pretty easy to convert to OXF. Perl is the best bet. Probably going to add a measure to reconcile the statement balance with the list of transactions so I make sure there's nothing dropped.