2007-01-21 GnuCash IRC logs
01:31:20 *** warlord is now known as warlord-afk
02:02:50 *** warlord-afk has quit IRC
03:28:56 *** hampton is now known as hampton|away
03:29:10 *** hampton2 has left #gnucash
03:32:22 *** wizkid238 has quit IRC
03:50:52 *** prock_ has joined #gnucash
03:59:36 *** prock has quit IRC
06:10:04 *** andi5 has joined #gnucash
06:10:05 *** gncbot sets mode: +o andi5
06:13:34 *** |gunni| has joined #gnucash
06:15:01 *** Def has quit IRC
06:39:23 *** Def has joined #gnucash
07:19:59 *** ceplma_ has quit IRC
09:59:08 *** prock_ is now known as prock
10:06:20 <prock> is there a way to reserve a nick on gimpnet?
10:06:20 <gncbot> prock: Sent 16 hours and 39 minutes ago: <jsled> the general form of the solution is add an explicit reference to the library containing the missing symbol to the link line for the module . Though I notice in src/backend/file/Makefile.am that there's a missing ${top_builddir} ... that might be it. I'm not sure from your message what dir the error was in.
10:38:47 *** MrN has joined #gnucash
10:47:29 <jsled> prock: not that I'm aware of; there's no basic services ... it's a pretty gimp-y irc network (HA HA HA!)
11:06:14 *** ErKa has joined #gnucash
11:08:51 * prock groans
11:09:53 *** warlord has joined #gnucash
11:09:54 *** gncbot sets mode: +o warlord
11:12:40 <jsled> morning warlord
11:51:19 *** andi5 has quit IRC
11:52:03 *** wizkid238 has joined #gnucash
11:54:16 *** andi5 has joined #gnucash
11:54:17 *** gncbot sets mode: +o andi5
11:59:50 <warlord> Thanks jsled
12:05:36 *** hampton|away is now known as hampton
12:07:43 *** mick0 has joined #gnucash
12:10:36 *** mick0 has quit IRC
12:18:56 <rearnsha> warlord: The change we discussed yesterday to the date formatting routines seems to have fixed all the crashes I was seeing. I'll file a bug tomorrow, when I can look up my gnome bugzilla account name (it's at work)
12:19:23 <rearnsha> Is this the right place to ask 'usage' questions?
12:19:46 <andi5> rearnsha: why do not you just click "forgot your password"? :)
12:20:09 <rearnsha> Because it would still send the info to my work email account :-(
12:20:25 <andi5> you cannot access your works email inbox?
12:20:35 <rearnsha> not trivially
12:20:41 <andi5> oook
12:21:07 <rearnsha> The patch, BTW is a one-liner.
12:21:21 <andi5> you need something applied?
12:21:26 <rearnsha> Yep
12:21:52 <rearnsha> BTW, I think it's probably more related to UK date formatting than NetBSD
12:21:59 <andi5> what is it?
12:22:05 <rearnsha> The patch?
12:22:07 <rearnsha> Hold on
12:22:11 <andi5> yes, please pastebin it
12:25:41 <rearnsha> andi5: http://pastebin.com/864369
12:25:54 <rearnsha> Watch out for the pastebin highlighting artifact
12:26:50 <andi5> too bad, aqbanking just stopped compiling... meaning additional work (more than just pressing enter)
12:27:16 <rearnsha> BTW: I didn't look too closely, but I thought I saw another potential MAX_DATE_LENGTH bug while I was grepping the sources. Might be worth a more general audit of this
12:30:21 <andi5> rearnsha: what is the related bugzilla entry?
12:30:55 <rearnsha> I haven't created it yet, 'cos my bugzilla account info is at work
12:31:09 <andi5> oh, i am sorry, i misread
12:31:29 <andi5> what happens?
12:32:13 <rearnsha> gnucash just crashes on almost any attempt to use transaction registers. Sometimes even simply opening one is enough
12:32:39 <rearnsha> The undersized buffer leads to stack corruption
12:33:01 <andi5> rearnsha: may you pastebin the stacktrace? maybe there are duplicates our there? ... what is your version, btw?
12:33:21 <rearnsha> http://pastebin.com/863733
12:33:22 <warlord> andi5: I went this with him yesterday
12:33:26 <warlord> It's really a bug.
12:33:31 <andi5> yes, i believe him
12:34:17 <rearnsha> version is 2.0.3
12:34:28 <andi5> oh, the logs... i did not read carefully enough
12:34:31 <rearnsha> but the SVN sources for trunk show the same code
12:36:09 <warlord> andi5: I already verified that this is an issue with him. No need to double-check. But I didn't see his patch.
12:36:37 <andi5> warlord: ok, shall i apply then?
12:37:32 <prock> are checkins double posted to gnucash-patches and gnucash-devel?
12:37:50 <warlord> I'm waiting for pastebin to show me the patch..
12:37:55 <warlord> prock: nope.
12:38:09 <warlord> but they ARE double posted to gnucash-patches and gnucash-commits.
12:38:16 <andi5> -patches and -changes, the latter the the full diff
12:42:06 <warlord> andi5: and yes, that patch looks fine
12:42:29 <andi5> that does not answer my Q: it is more about whether you or me ;-)
12:43:10 *** kielein has joined #gnucash
12:43:49 <andi5> ok, maybe it does...
12:49:16 <andi5> rearnsha: what is the name you want in the commit log?
12:49:39 <rearnsha> Richard Earnshaw <rearnsha@netbsd.org> is probably the best for this one.
12:50:03 *** mnoir has joined #gnucash
12:52:41 <andi5> rearnsha: thanks... r15409 is your ticket number :)
12:53:16 <rearnsha> Ta. Now all I need to do is find out why the Finance::Quote isn't working for me...
12:53:40 <rearnsha> "There was a system error" isn't very helpful :-(
12:55:09 <andi5> after having read parts of intltool-update i hate perl even more... i should add an emacs hook that kills the current buffer on a switch to perl-mode
12:56:25 <warlord> Heh
12:59:43 *** keyline has joined #gnucash
13:08:51 *** kielein has quit IRC
14:07:12 *** ErKa has quit IRC
14:13:54 <hampton> rearnsha: Try applying http://svn.gnucash.org/trac/changeset/15397
14:16:15 *** sjc has joined #gnucash
14:16:42 <rearnsha> Thanks, I'll take a look at that. My immediate problem was that I didn't have Date::Manip installed. The error message gave no clue to this problem and it was only when I ran gnc-fq-helper that the problem became apparent
14:17:12 <hampton> Yep. That's been fixed in trunk but not back-ported to 2.0 yet.
14:51:30 <jsled> Hmm. Any guesses what version of glib david reiser (on Mac OSX 10.4.8) might have, so I can figure out what gmem.c:154 is ?
14:54:02 <warlord> might need to ask him
14:54:16 * jsled just runs across the G_{,UN}LIKELY macros in glib.
14:55:36 <jsled> On the theory that there's not a whole lot of churn between glib-2.10 and 2.12, which are the most likely, the mess is in g_malloc0(...) which is used by probably pretty much everything.
14:56:02 <jsled> We don't do hardly any g_malloc()s ... only g_new0()s. And there's only a handful with a variable as the "size" arg.
14:56:26 <warlord> Huh
15:03:53 <jsled> warlord: when are you in CA, next?
15:04:49 <warlord> jsled: I dont know.
15:19:17 <prock> is there a doc describing how to launch gnucash from a non-standard prefix installation? It starts to launch and then gives me some strange schemish backtrace
15:20:46 <jsled> prock: ${prefix}/bin/gnucash should work.
15:21:00 <jsled> pastebin the backtrace?
15:21:10 <prock> why am I having so many troubles :)
15:23:21 <warlord> prock: what distro are you running?
15:23:25 <prock> Debian
15:23:28 <prock> (etch)
15:23:29 <warlord> Odd!
15:24:20 <prock> http://pastebin.com/864462
15:27:06 <warlord> That looks like an slib bug.
15:27:20 <warlord> Are you just using Debian's GnuCash packages?
15:27:32 <prock> no that's a trunk install
15:27:41 <jsled> but, otherwise?
15:27:49 <prock> yes
15:27:52 <warlord> Are you using Debian's slib?
15:28:01 <prock> 3a4-4
15:28:09 <jsled> bzzt.
15:28:10 <warlord> Huh
15:28:22 <prock> slib 3a4-4
15:28:26 <jsled> (or did debian get 3a4 working?)
15:29:06 <jsled> Alos, what guile and g-wrap?
15:29:58 <prock> guile 1.6.8-6 g-wrap 1.9.6-3.1
15:33:37 <jsled> gentoo has problems with guile-1.6.8 and >=slib-3.1.1, though it sounds like debian worked through them. Maybe they didn't, fully.
15:34:45 <jsled> Are those the same versions as your packaged gnucash-2.0.4 is using?
15:34:59 <jsled> And, what was your configure line, exactly?
15:35:44 <prock> ./configure --enable-compile-warnings --prefix=/home/pmcalpine/src/gnucash/install/trunk
15:36:29 <prock> yes that was what was installed to run the packaged gnucash... although the packaged gnucash is 2.0.2-2.1
15:36:39 <jsled> Weird!
15:37:31 <prock> the sx-cleanup branch runs ok.
15:37:38 <andi5> hehe
15:37:58 <warlord> heh.
15:47:03 <prock> what's the gbr_init error mean?
15:47:19 <warlord> You can ignore it.
15:54:55 <prock> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=405985
15:55:36 <andi5> Q: is one allowed to free/alter the return value of getenv()?
15:56:10 <prock> what's a distro that trunk actually works on?
15:56:37 <andi5> ubuntu, with downgraded slib, just one of them
15:56:52 <prock> which slib version?
15:57:24 <andi5> i do not know and cannot check ;-) ... 3a3 or so
16:02:59 <warlord> andi5: no.
16:03:13 <andi5> great
16:03:15 <warlord> prock: Works great for me on FC5
16:03:45 * andi5 is trying to trash all those funny encoding bugs on windows... g_getenv is the next ;)
16:04:47 <warlord> ISTR an issue with using g_gettenv()..
16:05:29 <andi5> yeah, one "t" too much ;-) .... any idea what it was?
16:06:24 <andi5> the alternative is to use getenv<->setenv as pair (but not getenv ->g_setenv)
16:07:16 *** CSWookie has joined #gnucash
16:07:35 <CSWookie> I just downloaded the new version of Gnucash, and I think I'm going to like it lots.
16:07:53 <warlord> Oh, I think the issue might've been when it was introduced into glib...
16:08:07 <CSWookie> Where by "new" I mean 4 months ago... :-)
16:08:10 <andi5> warlord: what about trying it out? we are on trunk anyway
16:08:33 <warlord> True. I think we can probably use it now.
16:08:45 <andi5> good :)
16:09:41 *** ceplma has joined #gnucash
16:17:42 *** sjc_ has joined #gnucash
16:26:34 *** sjc has quit IRC
16:29:57 <andi5> warlord: maybe setenv had issues? we only use g_setenv.... just an idea :)
16:31:08 <warlord> Maybe.
16:31:14 <warlord> You'd have to ask cstim
16:31:27 <andi5> ok
16:34:54 <andi5> warlord: yeah, i see two issues in trac: all setenv's have been replaced by g_setenv because of portability reasons... and then we needed to revert the meaning of the return value (0 = good, FALSE = bad)
16:35:10 <andi5> but well, we will see
16:35:28 <warlord> Okay
16:45:14 *** jpeach has joined #gnucash
16:45:58 *** jpeach has left #gnucash
16:50:25 <CSWookie> Is there any thought of adding undo-redo support?
16:53:43 <warlord> CSWookie: yes
16:53:53 <warlord> (thought, that is)
16:54:33 <CSWookie> warlord: Cool.
16:56:07 <andi5> hm.... i wonder whether that is right: stacktrace with 563 frames
17:05:52 <warlord> Hmm
17:15:57 <andi5> Q: is it possible to write a creat() as open() without loss of functionality?
17:16:16 <warlord> andi5: Yes.
17:16:20 <warlord> O_CREAT
17:16:32 <warlord> (with an optional O_EXCL)
17:17:01 <CSWookie> Hmm.
17:17:20 *** keyline has quit IRC
17:17:24 <CSWookie> So, my bank made an error, and gave me a 5$ credit on my credit card as part of the compensation.
17:17:29 <CSWookie> Where do I say that came from?
17:17:56 <andi5> i read: creat(path, mode) is equivalent to open(path, O_WRONLY | O_CREAT | O_TRUNC, mode) .... i will use that then, ok? (one case)
17:19:18 <andi5> but maybe it is the wrong platform.... hmpf
17:19:24 *** sjc__ has joined #gnucash
17:20:41 <andi5> may someone do a "man 2 creat" for me? ;-)
17:21:37 *** ceplma has quit IRC
17:22:05 <CSWookie> Evidently I don't have the C man pages on this box.
17:22:17 <CSWookie> Ubuntu is cracked out.
17:22:28 <andi5> CSWookie: apt-get install libc-doc or so
17:22:40 <CSWookie> It's vital that the desktop wallpaper be kept up to date, but C manpages are not very important.
17:22:40 <andi5> maybe glibc-doc
17:22:53 <andi5> that is consequency
17:22:55 <CSWookie> andi5: Yeah, I know.
17:24:07 <warlord> andi5: You may or may not want O_EXCL, depending on whether you want to fail if the file already exists.
17:24:35 <warlord> But the rest looks correct.
17:24:54 <andi5> (btw, it works... i can create files with names "öhööö" now)
17:25:45 <andi5> i am looking at http://svn.gnucash.org/trac/browser/gnucash/trunk/src/backend/file/gnc-backend-file.c#L350
17:26:10 <andi5> i guess i will need that, right?
17:26:33 <warlord> Probbaly, yeah.
17:26:35 <warlord> :)
17:27:28 *** sjc_ has quit IRC
17:29:35 <warlord> Man, this is annoying.... I'm trying to find the cost bases for an investment portfolio. They'll give me the cost bases for the entire portfolio but wont break it out on a per-fund basis within the portfolio.
17:31:48 <andi5> ok, so some crashes are left but will have to wait :) ... see you all... thanks warlord for your support
17:31:59 <warlord> see ya, andi5
17:32:04 *** andi5 has quit IRC
17:42:36 *** MrN has quit IRC
18:17:59 *** _gunni_ has joined #gnucash
18:27:05 *** |gunni| has quit IRC
18:31:53 <warlord> Man, that was a pain.
18:32:22 <rearnsha> Is there any way to get payments into a pension fund to appear as a tax-deductable expense?
18:35:09 <warlord> rearnsha: as opposed to what, in terms of gnucash?
18:35:41 <rearnsha> Well the tax report form only allows you to select income and expense accounts for reporting
18:36:24 <rearnsha> So the money comes out of income:salary and goes into an asset:pension category.
18:36:47 <rearnsha> but that amount of money is tax-deductable in the UK
18:38:04 <warlord> Um, the tax report is US-only.
18:38:09 <warlord> There is no UK tax report.
18:38:17 <warlord> (nobody has written one and donated it)
18:38:20 <rearnsha> :-(
18:38:32 <warlord> Actually, there's a German one, too..
18:38:35 <warlord> But that dosn't help you.
18:38:47 <warlord> Do you have the accounts appropriately marked "Tax Related"?
18:39:06 <rearnsha> I think so
18:39:08 * rearnsha checks
18:39:11 * warlord has never actually used the tax report(s)
18:40:58 <rearnsha> Yes, I've marked the pension folder as tax-related, but it doesn't appear in the report.
18:42:12 <warlord> I suggest you ask on -user.
19:32:19 *** wizkid238 has quit IRC
19:50:38 *** hampton2 has joined #gnucash
19:50:39 *** gncbot sets mode: +o hampton2
19:51:08 <warlord> hiya hampton
19:51:25 <hampton2> Hi warlord
19:52:11 <warlord> How are you doing?
19:54:09 <hampton2> Good. Keeping busy.
19:55:06 <hampton2> Trying to get the point where I can devote some time to gnucash.
19:55:27 <hampton2> Hey, can I get a review of 15269 and 15416? I'd like to get these into 2.0.
19:56:14 <hampton2> 15397 too.
19:56:20 <hampton2> They're all pretty simple changes.
19:59:15 <warlord> I'll try to take a look. There are a number of changes waiting for 2.0.. Plus one more I've got in private email that will need an audit.
20:00:38 <warlord> Right now I'm watching the NE/Indy game
20:03:02 <hampton2> Yeah, I've got it on too.
20:03:11 <warlord> well, trunk at least builds, completes its checks, and installs..
20:03:48 <warlord> I do think it would be appropriate to have a NE/Bears Superbowl. Reminiscent of 1984.
20:11:27 *** mnoir has quit IRC
20:34:15 *** _gunni_ has quit IRC
20:56:24 *** sjc__ has quit IRC
21:03:32 *** warlord is now known as warlord-afk
21:47:03 *** wizkid238 has joined #gnucash
21:49:25 *** swoolley has joined #gnucash
21:50:41 <swoolley> I'm testing gnucash 2 so that I can see if I'm ready to upgrade a 1.8 version, and all looks good except reports seem to not be respecting my preferences for plot width (although plot height is working) is this normal?
21:51:18 <swoolley> subversion revision r14741
21:56:38 <hampton2> Don't know about your exact question. The old graphing subsystem (guppi) was completely swapped out in 2.0 (for libgoffice).
21:56:46 <hampton2> (fyi)
21:58:03 <swoolley> hmm, well, I'll try to make sure libgoffice is reasonably up-to-date
22:01:22 <swoolley> the Building page on the wiki makes no mention of goffice, btw
22:02:08 <hampton2> gnucash includes a copy of libgoffice that it uses if your system libgoffice isn't new enough.
22:02:26 <swoolley> *cringe*
22:03:43 <hampton2> It was needed at the time because not all platforms shipped with libgoffice, or a recent enough version.
22:05:16 <swoolley> >= 0.3?
22:05:45 <swoolley> I had 0.2.1 installed
22:05:54 <swoolley> so it should have used its own ....
22:06:14 <hampton2> libgsf-1 >= 1.12.2
22:06:23 <hampton2> libgoffice-1 >= 0.0.4
22:06:52 <swoolley> Oh, I also see "libgoffice-0.3 >= 0.3.0"
22:08:34 <hampton2> I don't see that in the top level configure.in file.
22:08:56 <chris> swoolley: that plot width thing is a work-around for a bug in either gtkhtml or gog (I forget.)
22:09:52 <chris> swoolley: The _think_ the work-around will let widths work if a new enough version is detected.
22:10:04 <chris> s/The/I/
22:10:11 <swoolley> hampton2: hmm, I see it in my configure.in, right above the 0.0.4 check, but it doesn't look like it fails if that fails
22:11:15 <swoolley> chris: ahh, ok, maybe you can point me to the bug I can double-check to see exactly what I need
22:11:19 <hampton2> Ah, you're running trunk, not 2.0.x.
22:11:41 <swoolley> yeah, I'm running straight out of the trunk.
22:12:00 <swoolley> I can pull a regular tag if that's better :)
22:12:09 <swoolley> 2.0.4?
22:13:10 <swoolley> and where are the scheduled transactions stored? Those didn't seem to come over automatically
22:13:30 <hampton2> your choice. Head is seeing some churn in the last week or so. branches/2.0 is what I was looking at.
22:13:33 <swoolley> I just copied my gnucash.gz file over to a testing box
22:14:44 <hampton2> I thought those were in the data file. Check your prefs. I think the default setting as to whether to run then at startup may have gotten inverted in 2.0
22:16:01 <swoolley> I went and looked in the table for them and there were none.
22:17:14 <swoolley> although I just uninstalled it so I can install a new version
22:17:44 <hampton2> that's odd. the more that I think about it, the more positive I am that the scheduled transactions are part of the data file.
22:18:03 <swoolley> I'll try it again when it finishes building
22:18:41 <chris> SXs are in the datafile.
22:18:51 <chris> swoolley: are all your graphs square?
22:19:00 <swoolley> chris: they are all square
22:19:16 <swoolley> change the size larger, the width goes the same width as it is high
22:19:27 <swoolley> linked to the height
22:19:40 <chris> [[
22:19:42 <swoolley> I'm upgrading gtkhtml right now
22:19:42 <chris> static void
22:19:42 <chris> gtkhtml_pre_3_10_1_bug_workaround(GtkHTMLEmbedded *eb)
22:19:42 <chris> {
22:19:42 <chris> /* HACK ALERT! Compensate for bug in gtkhtml < 3.10.1
22:19:42 <chris> Gtkhtml set the width parameter twice (=width, =height), so both,
22:19:45 <chris> width (==height) and height (<1) were incorrect. */
22:19:47 <chris> ]]
22:20:07 <swoolley> I was using 3.8, upgrading to 3.12
22:20:47 <swoolley> chris: nice
22:21:55 <swoolley> chris: well, I can confirm that the workaround is doing its job :)
22:22:09 <chris> unfortunately, a quick read of the code suggests that the work-around (squaring the graphs) is always applied. i.e. we don't actually detect the gtk_html version.
22:23:01 *** twunder has joined #gnucash
22:23:02 <chris> feel free to submit a patch that defines the appropriate symbol in configure.in and then uses it in gnc-html-graph-gog.c.
22:23:05 <swoolley> it might do the check up the function stack
22:23:35 <swoolley> yeah, I'll neuterize it if it still doesn't work after this rebuild
22:24:15 <swoolley> chris: you want a patch against HEAD or 2.0.4?
22:25:27 <swoolley> I'll assume HEAD
22:26:31 <hampton2> against trunk, please.
22:27:53 <swoolley> ok
22:56:08 *** twunder has quit IRC
22:58:35 *** warlord-afk is now known as warlord
23:01:29 <swoolley> chris: looking at the code, the workaround only takes effect if the height value is wrong, so I shouldn't need to patch anything
23:04:02 *** hampton2 has quit IRC
23:04:20 <chris> swoolley: oh.. yeah.. I remember now...
23:05:01 <chris> So it should work with the newer gtkhtml.
23:11:37 <swoolley> chris: :)
23:13:01 <swoolley> thanks anyways, guys.
23:27:55 *** wizkid238 has quit IRC
23:32:29 *** wizkid238 has joined #gnucash
23:36:27 <chris> swoolley: So, does it work with newer gtkhtml?
01:31:20 *** warlord is now known as warlord-afk
02:02:50 *** warlord-afk has quit IRC
03:28:56 *** hampton is now known as hampton|away
03:29:10 *** hampton2 has left #gnucash
03:32:22 *** wizkid238 has quit IRC
03:50:52 *** prock_ has joined #gnucash
03:59:36 *** prock has quit IRC
06:10:04 *** andi5 has joined #gnucash
06:10:05 *** gncbot sets mode: +o andi5
06:13:34 *** |gunni| has joined #gnucash
06:15:01 *** Def has quit IRC
06:39:23 *** Def has joined #gnucash
07:19:59 *** ceplma_ has quit IRC
09:59:08 *** prock_ is now known as prock
10:06:20 <prock> is there a way to reserve a nick on gimpnet?
10:06:20 <gncbot> prock: Sent 16 hours and 39 minutes ago: <jsled> the general form of the solution is add an explicit reference to the library containing the missing symbol to the link line for the module . Though I notice in src/backend/file/Makefile.am that there's a missing ${top_builddir} ... that might be it. I'm not sure from your message what dir the error was in.
10:38:47 *** MrN has joined #gnucash
10:47:29 <jsled> prock: not that I'm aware of; there's no basic services ... it's a pretty gimp-y irc network (HA HA HA!)
11:06:14 *** ErKa has joined #gnucash
11:08:51 * prock groans
11:09:53 *** warlord has joined #gnucash
11:09:54 *** gncbot sets mode: +o warlord
11:12:40 <jsled> morning warlord
11:51:19 *** andi5 has quit IRC
11:52:03 *** wizkid238 has joined #gnucash
11:54:16 *** andi5 has joined #gnucash
11:54:17 *** gncbot sets mode: +o andi5
11:59:50 <warlord> Thanks jsled
12:05:36 *** hampton|away is now known as hampton
12:07:43 *** mick0 has joined #gnucash
12:10:36 *** mick0 has quit IRC
12:18:56 <rearnsha> warlord: The change we discussed yesterday to the date formatting routines seems to have fixed all the crashes I was seeing. I'll file a bug tomorrow, when I can look up my gnome bugzilla account name (it's at work)
12:19:23 <rearnsha> Is this the right place to ask 'usage' questions?
12:19:46 <andi5> rearnsha: why do not you just click "forgot your password"? :)
12:20:09 <rearnsha> Because it would still send the info to my work email account :-(
12:20:25 <andi5> you cannot access your works email inbox?
12:20:35 <rearnsha> not trivially
12:20:41 <andi5> oook
12:21:07 <rearnsha> The patch, BTW is a one-liner.
12:21:21 <andi5> you need something applied?
12:21:26 <rearnsha> Yep
12:21:52 <rearnsha> BTW, I think it's probably more related to UK date formatting than NetBSD
12:21:59 <andi5> what is it?
12:22:05 <rearnsha> The patch?
12:22:07 <rearnsha> Hold on
12:22:11 <andi5> yes, please pastebin it
12:25:41 <rearnsha> andi5: http://pastebin.com/864369
12:25:54 <rearnsha> Watch out for the pastebin highlighting artifact
12:26:50 <andi5> too bad, aqbanking just stopped compiling... meaning additional work (more than just pressing enter)
12:27:16 <rearnsha> BTW: I didn't look too closely, but I thought I saw another potential MAX_DATE_LENGTH bug while I was grepping the sources. Might be worth a more general audit of this
12:30:21 <andi5> rearnsha: what is the related bugzilla entry?
12:30:55 <rearnsha> I haven't created it yet, 'cos my bugzilla account info is at work
12:31:09 <andi5> oh, i am sorry, i misread
12:31:29 <andi5> what happens?
12:32:13 <rearnsha> gnucash just crashes on almost any attempt to use transaction registers. Sometimes even simply opening one is enough
12:32:39 <rearnsha> The undersized buffer leads to stack corruption
12:33:01 <andi5> rearnsha: may you pastebin the stacktrace? maybe there are duplicates our there? ... what is your version, btw?
12:33:21 <rearnsha> http://pastebin.com/863733
12:33:22 <warlord> andi5: I went this with him yesterday
12:33:26 <warlord> It's really a bug.
12:33:31 <andi5> yes, i believe him
12:34:17 <rearnsha> version is 2.0.3
12:34:28 <andi5> oh, the logs... i did not read carefully enough
12:34:31 <rearnsha> but the SVN sources for trunk show the same code
12:36:09 <warlord> andi5: I already verified that this is an issue with him. No need to double-check. But I didn't see his patch.
12:36:37 <andi5> warlord: ok, shall i apply then?
12:37:32 <prock> are checkins double posted to gnucash-patches and gnucash-devel?
12:37:50 <warlord> I'm waiting for pastebin to show me the patch..
12:37:55 <warlord> prock: nope.
12:38:09 <warlord> but they ARE double posted to gnucash-patches and gnucash-commits.
12:38:16 <andi5> -patches and -changes, the latter the the full diff
12:42:06 <warlord> andi5: and yes, that patch looks fine
12:42:29 <andi5> that does not answer my Q: it is more about whether you or me ;-)
12:43:10 *** kielein has joined #gnucash
12:43:49 <andi5> ok, maybe it does...
12:49:16 <andi5> rearnsha: what is the name you want in the commit log?
12:49:39 <rearnsha> Richard Earnshaw <rearnsha@netbsd.org> is probably the best for this one.
12:50:03 *** mnoir has joined #gnucash
12:52:41 <andi5> rearnsha: thanks... r15409 is your ticket number :)
12:53:16 <rearnsha> Ta. Now all I need to do is find out why the Finance::Quote isn't working for me...
12:53:40 <rearnsha> "There was a system error" isn't very helpful :-(
12:55:09 <andi5> after having read parts of intltool-update i hate perl even more... i should add an emacs hook that kills the current buffer on a switch to perl-mode
12:56:25 <warlord> Heh
12:59:43 *** keyline has joined #gnucash
13:08:51 *** kielein has quit IRC
14:07:12 *** ErKa has quit IRC
14:13:54 <hampton> rearnsha: Try applying http://svn.gnucash.org/trac/changeset/15397
14:16:15 *** sjc has joined #gnucash
14:16:42 <rearnsha> Thanks, I'll take a look at that. My immediate problem was that I didn't have Date::Manip installed. The error message gave no clue to this problem and it was only when I ran gnc-fq-helper that the problem became apparent
14:17:12 <hampton> Yep. That's been fixed in trunk but not back-ported to 2.0 yet.
14:51:30 <jsled> Hmm. Any guesses what version of glib david reiser (on Mac OSX 10.4.8) might have, so I can figure out what gmem.c:154 is ?
14:54:02 <warlord> might need to ask him
14:54:16 * jsled just runs across the G_{,UN}LIKELY macros in glib.
14:55:36 <jsled> On the theory that there's not a whole lot of churn between glib-2.10 and 2.12, which are the most likely, the mess is in g_malloc0(...) which is used by probably pretty much everything.
14:56:02 <jsled> We don't do hardly any g_malloc()s ... only g_new0()s. And there's only a handful with a variable as the "size" arg.
14:56:26 <warlord> Huh
15:03:53 <jsled> warlord: when are you in CA, next?
15:04:49 <warlord> jsled: I dont know.
15:19:17 <prock> is there a doc describing how to launch gnucash from a non-standard prefix installation? It starts to launch and then gives me some strange schemish backtrace
15:20:46 <jsled> prock: ${prefix}/bin/gnucash should work.
15:21:00 <jsled> pastebin the backtrace?
15:21:10 <prock> why am I having so many troubles :)
15:23:21 <warlord> prock: what distro are you running?
15:23:25 <prock> Debian
15:23:28 <prock> (etch)
15:23:29 <warlord> Odd!
15:24:20 <prock> http://pastebin.com/864462
15:27:06 <warlord> That looks like an slib bug.
15:27:20 <warlord> Are you just using Debian's GnuCash packages?
15:27:32 <prock> no that's a trunk install
15:27:41 <jsled> but, otherwise?
15:27:49 <prock> yes
15:27:52 <warlord> Are you using Debian's slib?
15:28:01 <prock> 3a4-4
15:28:09 <jsled> bzzt.
15:28:10 <warlord> Huh
15:28:22 <prock> slib 3a4-4
15:28:26 <jsled> (or did debian get 3a4 working?)
15:29:06 <jsled> Alos, what guile and g-wrap?
15:29:58 <prock> guile 1.6.8-6 g-wrap 1.9.6-3.1
15:33:37 <jsled> gentoo has problems with guile-1.6.8 and >=slib-3.1.1, though it sounds like debian worked through them. Maybe they didn't, fully.
15:34:45 <jsled> Are those the same versions as your packaged gnucash-2.0.4 is using?
15:34:59 <jsled> And, what was your configure line, exactly?
15:35:44 <prock> ./configure --enable-compile-warnings --prefix=/home/pmcalpine/src/gnucash/install/trunk
15:36:29 <prock> yes that was what was installed to run the packaged gnucash... although the packaged gnucash is 2.0.2-2.1
15:36:39 <jsled> Weird!
15:37:31 <prock> the sx-cleanup branch runs ok.
15:37:38 <andi5> hehe
15:37:58 <warlord> heh.
15:47:03 <prock> what's the gbr_init error mean?
15:47:19 <warlord> You can ignore it.
15:54:55 <prock> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=405985
15:55:36 <andi5> Q: is one allowed to free/alter the return value of getenv()?
15:56:10 <prock> what's a distro that trunk actually works on?
15:56:37 <andi5> ubuntu, with downgraded slib, just one of them
15:56:52 <prock> which slib version?
15:57:24 <andi5> i do not know and cannot check ;-) ... 3a3 or so
16:02:59 <warlord> andi5: no.
16:03:13 <andi5> great
16:03:15 <warlord> prock: Works great for me on FC5
16:03:45 * andi5 is trying to trash all those funny encoding bugs on windows... g_getenv is the next ;)
16:04:47 <warlord> ISTR an issue with using g_gettenv()..
16:05:29 <andi5> yeah, one "t" too much ;-) .... any idea what it was?
16:06:24 <andi5> the alternative is to use getenv<->setenv as pair (but not getenv ->g_setenv)
16:07:16 *** CSWookie has joined #gnucash
16:07:35 <CSWookie> I just downloaded the new version of Gnucash, and I think I'm going to like it lots.
16:07:53 <warlord> Oh, I think the issue might've been when it was introduced into glib...
16:08:07 <CSWookie> Where by "new" I mean 4 months ago... :-)
16:08:10 <andi5> warlord: what about trying it out? we are on trunk anyway
16:08:33 <warlord> True. I think we can probably use it now.
16:08:45 <andi5> good :)
16:09:41 *** ceplma has joined #gnucash
16:17:42 *** sjc_ has joined #gnucash
16:26:34 *** sjc has quit IRC
16:29:57 <andi5> warlord: maybe setenv had issues? we only use g_setenv.... just an idea :)
16:31:08 <warlord> Maybe.
16:31:14 <warlord> You'd have to ask cstim
16:31:27 <andi5> ok
16:34:54 <andi5> warlord: yeah, i see two issues in trac: all setenv's have been replaced by g_setenv because of portability reasons... and then we needed to revert the meaning of the return value (0 = good, FALSE = bad)
16:35:10 <andi5> but well, we will see
16:35:28 <warlord> Okay
16:45:14 *** jpeach has joined #gnucash
16:45:58 *** jpeach has left #gnucash
16:50:25 <CSWookie> Is there any thought of adding undo-redo support?
16:53:43 <warlord> CSWookie: yes
16:53:53 <warlord> (thought, that is)
16:54:33 <CSWookie> warlord: Cool.
16:56:07 <andi5> hm.... i wonder whether that is right: stacktrace with 563 frames
17:05:52 <warlord> Hmm
17:15:57 <andi5> Q: is it possible to write a creat() as open() without loss of functionality?
17:16:16 <warlord> andi5: Yes.
17:16:20 <warlord> O_CREAT
17:16:32 <warlord> (with an optional O_EXCL)
17:17:01 <CSWookie> Hmm.
17:17:20 *** keyline has quit IRC
17:17:24 <CSWookie> So, my bank made an error, and gave me a 5$ credit on my credit card as part of the compensation.
17:17:29 <CSWookie> Where do I say that came from?
17:17:56 <andi5> i read: creat(path, mode) is equivalent to open(path, O_WRONLY | O_CREAT | O_TRUNC, mode) .... i will use that then, ok? (one case)
17:19:18 <andi5> but maybe it is the wrong platform.... hmpf
17:19:24 *** sjc__ has joined #gnucash
17:20:41 <andi5> may someone do a "man 2 creat" for me? ;-)
17:21:37 *** ceplma has quit IRC
17:22:05 <CSWookie> Evidently I don't have the C man pages on this box.
17:22:17 <CSWookie> Ubuntu is cracked out.
17:22:28 <andi5> CSWookie: apt-get install libc-doc or so
17:22:40 <CSWookie> It's vital that the desktop wallpaper be kept up to date, but C manpages are not very important.
17:22:40 <andi5> maybe glibc-doc
17:22:53 <andi5> that is consequency
17:22:55 <CSWookie> andi5: Yeah, I know.
17:24:07 <warlord> andi5: You may or may not want O_EXCL, depending on whether you want to fail if the file already exists.
17:24:35 <warlord> But the rest looks correct.
17:24:54 <andi5> (btw, it works... i can create files with names "öhööö" now)
17:25:45 <andi5> i am looking at http://svn.gnucash.org/trac/browser/gnucash/trunk/src/backend/file/gnc-backend-file.c#L350
17:26:10 <andi5> i guess i will need that, right?
17:26:33 <warlord> Probbaly, yeah.
17:26:35 <warlord> :)
17:27:28 *** sjc_ has quit IRC
17:29:35 <warlord> Man, this is annoying.... I'm trying to find the cost bases for an investment portfolio. They'll give me the cost bases for the entire portfolio but wont break it out on a per-fund basis within the portfolio.
17:31:48 <andi5> ok, so some crashes are left but will have to wait :) ... see you all... thanks warlord for your support
17:31:59 <warlord> see ya, andi5
17:32:04 *** andi5 has quit IRC
17:42:36 *** MrN has quit IRC
18:17:59 *** _gunni_ has joined #gnucash
18:27:05 *** |gunni| has quit IRC
18:31:53 <warlord> Man, that was a pain.
18:32:22 <rearnsha> Is there any way to get payments into a pension fund to appear as a tax-deductable expense?
18:35:09 <warlord> rearnsha: as opposed to what, in terms of gnucash?
18:35:41 <rearnsha> Well the tax report form only allows you to select income and expense accounts for reporting
18:36:24 <rearnsha> So the money comes out of income:salary and goes into an asset:pension category.
18:36:47 <rearnsha> but that amount of money is tax-deductable in the UK
18:38:04 <warlord> Um, the tax report is US-only.
18:38:09 <warlord> There is no UK tax report.
18:38:17 <warlord> (nobody has written one and donated it)
18:38:20 <rearnsha> :-(
18:38:32 <warlord> Actually, there's a German one, too..
18:38:35 <warlord> But that dosn't help you.
18:38:47 <warlord> Do you have the accounts appropriately marked "Tax Related"?
18:39:06 <rearnsha> I think so
18:39:08 * rearnsha checks
18:39:11 * warlord has never actually used the tax report(s)
18:40:58 <rearnsha> Yes, I've marked the pension folder as tax-related, but it doesn't appear in the report.
18:42:12 <warlord> I suggest you ask on -user.
19:32:19 *** wizkid238 has quit IRC
19:50:38 *** hampton2 has joined #gnucash
19:50:39 *** gncbot sets mode: +o hampton2
19:51:08 <warlord> hiya hampton
19:51:25 <hampton2> Hi warlord
19:52:11 <warlord> How are you doing?
19:54:09 <hampton2> Good. Keeping busy.
19:55:06 <hampton2> Trying to get the point where I can devote some time to gnucash.
19:55:27 <hampton2> Hey, can I get a review of 15269 and 15416? I'd like to get these into 2.0.
19:56:14 <hampton2> 15397 too.
19:56:20 <hampton2> They're all pretty simple changes.
19:59:15 <warlord> I'll try to take a look. There are a number of changes waiting for 2.0.. Plus one more I've got in private email that will need an audit.
20:00:38 <warlord> Right now I'm watching the NE/Indy game
20:03:02 <hampton2> Yeah, I've got it on too.
20:03:11 <warlord> well, trunk at least builds, completes its checks, and installs..
20:03:48 <warlord> I do think it would be appropriate to have a NE/Bears Superbowl. Reminiscent of 1984.
20:11:27 *** mnoir has quit IRC
20:34:15 *** _gunni_ has quit IRC
20:56:24 *** sjc__ has quit IRC
21:03:32 *** warlord is now known as warlord-afk
21:47:03 *** wizkid238 has joined #gnucash
21:49:25 *** swoolley has joined #gnucash
21:50:41 <swoolley> I'm testing gnucash 2 so that I can see if I'm ready to upgrade a 1.8 version, and all looks good except reports seem to not be respecting my preferences for plot width (although plot height is working) is this normal?
21:51:18 <swoolley> subversion revision r14741
21:56:38 <hampton2> Don't know about your exact question. The old graphing subsystem (guppi) was completely swapped out in 2.0 (for libgoffice).
21:56:46 <hampton2> (fyi)
21:58:03 <swoolley> hmm, well, I'll try to make sure libgoffice is reasonably up-to-date
22:01:22 <swoolley> the Building page on the wiki makes no mention of goffice, btw
22:02:08 <hampton2> gnucash includes a copy of libgoffice that it uses if your system libgoffice isn't new enough.
22:02:26 <swoolley> *cringe*
22:03:43 <hampton2> It was needed at the time because not all platforms shipped with libgoffice, or a recent enough version.
22:05:16 <swoolley> >= 0.3?
22:05:45 <swoolley> I had 0.2.1 installed
22:05:54 <swoolley> so it should have used its own ....
22:06:14 <hampton2> libgsf-1 >= 1.12.2
22:06:23 <hampton2> libgoffice-1 >= 0.0.4
22:06:52 <swoolley> Oh, I also see "libgoffice-0.3 >= 0.3.0"
22:08:34 <hampton2> I don't see that in the top level configure.in file.
22:08:56 <chris> swoolley: that plot width thing is a work-around for a bug in either gtkhtml or gog (I forget.)
22:09:52 <chris> swoolley: The _think_ the work-around will let widths work if a new enough version is detected.
22:10:04 <chris> s/The/I/
22:10:11 <swoolley> hampton2: hmm, I see it in my configure.in, right above the 0.0.4 check, but it doesn't look like it fails if that fails
22:11:15 <swoolley> chris: ahh, ok, maybe you can point me to the bug I can double-check to see exactly what I need
22:11:19 <hampton2> Ah, you're running trunk, not 2.0.x.
22:11:41 <swoolley> yeah, I'm running straight out of the trunk.
22:12:00 <swoolley> I can pull a regular tag if that's better :)
22:12:09 <swoolley> 2.0.4?
22:13:10 <swoolley> and where are the scheduled transactions stored? Those didn't seem to come over automatically
22:13:30 <hampton2> your choice. Head is seeing some churn in the last week or so. branches/2.0 is what I was looking at.
22:13:33 <swoolley> I just copied my gnucash.gz file over to a testing box
22:14:44 <hampton2> I thought those were in the data file. Check your prefs. I think the default setting as to whether to run then at startup may have gotten inverted in 2.0
22:16:01 <swoolley> I went and looked in the table for them and there were none.
22:17:14 <swoolley> although I just uninstalled it so I can install a new version
22:17:44 <hampton2> that's odd. the more that I think about it, the more positive I am that the scheduled transactions are part of the data file.
22:18:03 <swoolley> I'll try it again when it finishes building
22:18:41 <chris> SXs are in the datafile.
22:18:51 <chris> swoolley: are all your graphs square?
22:19:00 <swoolley> chris: they are all square
22:19:16 <swoolley> change the size larger, the width goes the same width as it is high
22:19:27 <swoolley> linked to the height
22:19:40 <chris> [[
22:19:42 <swoolley> I'm upgrading gtkhtml right now
22:19:42 <chris> static void
22:19:42 <chris> gtkhtml_pre_3_10_1_bug_workaround(GtkHTMLEmbedded *eb)
22:19:42 <chris> {
22:19:42 <chris> /* HACK ALERT! Compensate for bug in gtkhtml < 3.10.1
22:19:42 <chris> Gtkhtml set the width parameter twice (=width, =height), so both,
22:19:45 <chris> width (==height) and height (<1) were incorrect. */
22:19:47 <chris> ]]
22:20:07 <swoolley> I was using 3.8, upgrading to 3.12
22:20:47 <swoolley> chris: nice
22:21:55 <swoolley> chris: well, I can confirm that the workaround is doing its job :)
22:22:09 <chris> unfortunately, a quick read of the code suggests that the work-around (squaring the graphs) is always applied. i.e. we don't actually detect the gtk_html version.
22:23:01 *** twunder has joined #gnucash
22:23:02 <chris> feel free to submit a patch that defines the appropriate symbol in configure.in and then uses it in gnc-html-graph-gog.c.
22:23:05 <swoolley> it might do the check up the function stack
22:23:35 <swoolley> yeah, I'll neuterize it if it still doesn't work after this rebuild
22:24:15 <swoolley> chris: you want a patch against HEAD or 2.0.4?
22:25:27 <swoolley> I'll assume HEAD
22:26:31 <hampton2> against trunk, please.
22:27:53 <swoolley> ok
22:56:08 *** twunder has quit IRC
22:58:35 *** warlord-afk is now known as warlord
23:01:29 <swoolley> chris: looking at the code, the workaround only takes effect if the height value is wrong, so I shouldn't need to patch anything
23:04:02 *** hampton2 has quit IRC
23:04:20 <chris> swoolley: oh.. yeah.. I remember now...
23:05:01 <chris> So it should work with the newer gtkhtml.
23:11:37 <swoolley> chris: :)
23:13:01 <swoolley> thanks anyways, guys.
23:27:55 *** wizkid238 has quit IRC
23:32:29 *** wizkid238 has joined #gnucash
23:36:27 <chris> swoolley: So, does it work with newer gtkhtml?