2008-09-05 GnuCash IRC logs
01:46:04 *** esperegu has joined #gnucash
03:58:40 *** ben_goodger has quit IRC
04:06:34 *** ben_goodger has joined #gnucash
04:19:53 *** ben_goodger has quit IRC
04:20:35 *** ben_goodger has joined #gnucash
04:55:39 *** _gunni_ has joined #gnucash
05:03:44 *** |gunni| has quit IRC
06:44:40 *** andi5 has joined #gnucash
06:44:41 *** gncbot sets mode: +o andi5
07:08:03 *** Pupeno has joined #gnucash
07:08:21 <Pupeno> Hello.
07:27:22 *** twunder has joined #gnucash
07:33:46 *** JimRaehl has left #gnucash
07:40:33 *** twunder has quit IRC
07:43:19 *** JimRaehl has joined #gnucash
08:19:42 *** Pupeno has left #gnucash
09:30:32 *** warlord-afk is now known as warlord
09:52:16 *** martincleaver has quit IRC
10:03:31 *** Mike1 has joined #gnucash
10:03:39 *** Mike1 has left #gnucash
10:28:09 *** martincleaver has joined #gnucash
10:35:02 *** martincleaver_ has joined #gnucash
10:43:38 *** martincleaver has quit IRC
12:03:45 *** Charles has joined #gnucash
12:06:14 <Charles> Are there problems with the mail server? My comments from last night's SVN commit (r17493) seem to have had trouble reaching gnucash-changes@gnucash.org and gnucash-patches@gnucash.org. Plus I see no new mail today.
12:07:08 <warlord> There were. The problems have been corrected and mail IS flowing now.
12:07:22 <warlord> it'll just take a while for everything to catch up.
12:07:28 <warlord> (the clamav milter crashed)
12:07:51 <Charles> Ah, that thing again. OK, thanks.
12:07:53 <warlord> (hi Charles -- you should hang out here more often!)
12:08:00 *** andrewsw-afk is now known as andrewsw
12:08:24 <warlord> Yeah... I should consider updating the server to F9... Or maybe I'll wait a few months and update to F10. *sighs*
12:09:02 <andrewsw> any movement on the new server front? oh and what was the problem while you were in (somewhere)
12:09:22 <warlord> No movement on the new server.
12:09:30 <warlord> I have no idea what the problem was -- it hasn't happened again.
12:09:41 <andrewsw> neutrinos?
12:10:07 <Charles> Chatting is OK, but I prefer email. I get interrupted a lot so I would have trouble responding in real time.
12:11:01 <warlord> we all get interrupted a lot ... but sometimes (many times) it's better to have an interactive conversation when you can swap-in a few minutes.. it can be much faster than going back and forth via email.
12:11:13 <warlord> (I certainly go idle for hours at a time)
12:11:22 <andrewsw> that's why the topic is what it is... lots of people are only partially here, even when they're here.
12:13:01 <Charles> I can could sit in here, but often unable to respond
12:14:47 <warlord> that's what lots of us do.. but at least when you have the time you could peek in ;)
12:15:33 <Charles> Just received my first gnucash email
12:18:07 <warlord> hmm.. I got the r17493 emails shortly after I kicked it.. *shrugs* same day service in a nanosecond world?
12:18:29 <Charles> If anyone wants to test r17493, I'd appreciate it. It works great for me on vanilla XP and XP/mingw. But since it isn't a small change and it involves the GNCDateEdit widget (again), the more testing, the better.
12:19:53 <Charles> No changes to the date/time calculations though. Just GTK changes.
12:20:10 <andrewsw> hmmm... windows. I haven't played with windows in a while. I might have to fire up my qemu instance...
12:20:41 <warlord> I'll try to test it on Linux at some point.
12:22:15 <Charles> I wasn't sure what to do with all those bugzilla reports related to that fix. Three are the same underlying problem with different symptoms. In general, would you combine those (like I did with the previous GNCDateEdit problem) or keep them separate?
12:22:38 <andrewsw> "the hardware on the computer has changed significantly... Windows must be reactivated within 3 days..."
12:22:49 <andrewsw> huh. not likely.
12:22:59 <warlord> If it's the same underlying problem with three symptoms you might designate one as the master and then have the others depend on it.
12:23:37 <Charles> I just made all three depend on the backport of the one fix (r17493)
12:24:39 <warlord> that works.
12:28:27 <Charles> OK. I think in future I will go with your idea, i.e. mark the earliest report as the master, then tell all newer reports (of the same underlying problem, but different symptoms) to use the master for all additional comments/updates. That way people have a better chance of seeing a duplicate report before they file, because there are still a variety of symptoms to match on.
12:31:22 *** Ozaed_ has joined #gnucash
12:32:03 <warlord> right
12:32:31 <warlord> You could just make it 'depend on' the earlier bug.. and then you can close them all at once when the underlying problem is fixed.
12:36:24 <Charles> Sounds good. Just marked another duplicate of 545722... will handle that with the 'depend' option in future.
12:39:06 <warlord> Well, sometimes it's easier to mark it as a dup.. ;)
12:40:07 *** Ozaed has quit IRC
12:47:16 <andrewsw> Charles: is that your progress bar in the qif import? nice!
12:49:56 <Charles> Yes, thanks. It was a pain! But the good news is that a lot of the work was extensions to the generic progress dialog routines, so that kind of display can be done anywhere now (not just in QIF). See gnome/dialog-progress.h
12:51:30 <andrewsw> I finally got around to updating my working system and saw that. Nice. thanks for it.
13:27:05 *** Mike1 has joined #gnucash
13:27:11 *** Mike1 has left #gnucash
13:47:16 <andrewsw> is it a side-effect of being able to watch the progress bar, is it actually running slower than previously?
13:47:39 <andrewsw> (purely anecdotal)
13:50:03 <warlord> Well, it IS doing more.. And probably more context switching between scheme and C
13:51:02 <andrewsw> yeah, makes sense I guess. I expect it to take a while to find dupes, but it took a good bit of time to "convert" the qif file as well.
13:51:45 <warlord> "convert"?
13:52:16 <andrewsw> one of the steps shown in the progress bar dialog is "converting /path/to/qif_file" or somesuch
13:52:35 <andrewsw> I guess that would be converting the qif into actual gnc txns.
13:58:16 <andrewsw> about 1 minute on "convert" and about 40 seconds on finding dupes.
13:58:35 <warlord> Could be. Or it could be that there are other things going on that fail to update the progress bar.
13:59:39 <andrewsw> sure
14:00:04 *** martincleaver_ has quit IRC
14:00:14 *** Zoolooc_ has joined #gnucash
14:00:31 <andrewsw> I think I'm probably an edge case for time complexity though at 1.9M gzipped.
14:04:39 <Charles> Andrew: The QIF importer speed has actually improved, I believe. My 2.5MB qif file now parses about 10x faster than a few versions ago.
14:05:39 <Charles> In previous versions, the progress bar got updated way too often. Now it gets updated less frequently, so it runs faster. Only marginally so on the conversion part, unfortunately.
14:06:06 <Charles> (When I say previous versions, I mean versions that had the "popup" progress bar.)
14:06:39 <andrewsw> interesting. I'm pretty sure it's slowed down significantly since 2.2.4, but that didn't have a progress bar at all, that I noticed.
14:07:17 <andrewsw> I wonder if it has to do with the number of splits, my qif import is a set of payroll transactions with about 15 splits
14:07:38 <andrewsw> per txn and around 15 txns
14:08:36 <andrewsw> oh, my bad. You're talking about a 2.5M qif file. My gnc file is 1.9M gzipped, so I expected finding dupes to be more time-consuming than the conversion.
14:08:45 <andrewsw> (that assumes that it's actually "Converting")
14:09:24 <andrewsw> my qif file is only 13k.
14:10:31 <Charles> I don't think the duplicate checking routines have changed recently. But previously the importer showed a separate progress bar during the duplicate checking phase. Now the "conversion" and "duplicate checking" phases are combined into one progress bar.
14:11:29 <Charles> Anyway if it is slower, I don't think it would be progress-bar-related.
14:12:11 <andrewsw> I'm not supposing it's progress bar related, just suggesting that perhaps the progress bar has made me more aware of it...
14:12:21 <andrewsw> apparent slowness may be a side-effect of that awareness.
14:12:52 <andrewsw> dang my gnc file is getting big. 21M uncompressed
14:13:22 <warlord> Wow. Mine is still only 250KB compressed.
14:13:35 <Charles> Actually, I do recall that there were some duplicate checking changes specifically related to split transactions...
14:13:36 <warlord> (but it only has data from 2007-01-01)
14:13:38 *** gantrixx has quit IRC
14:19:49 <andrewsw> mine is all company data from sometime in 2004. I really should use the book closing feature...
14:21:18 <warlord> that doesn't reduce the data.
14:21:26 <warlord> it just zero-izes the Income/Expense accounts.
14:21:54 <warlord> (and might cause some reports to not work right)
14:23:45 <Charles> Closing your books probably wouldn't speed up the duplicate checking. In fact, it might show it down. (more txns, more to check for duplication)
14:26:43 <warlord> Right.
14:38:11 <andrewsw> oh yeah, I guess I knew that.
14:39:46 *** martincleaver has joined #gnucash
14:54:09 *** Charles has quit IRC
14:57:17 <andrewsw> oh wow... segfault.
14:58:26 <andrewsw> Backtrace: In unknown file: ?: 0* Segmentation fault
14:59:22 <andrewsw> autosave to the rescue...
15:00:10 <warlord> neat
15:00:30 <andrewsw> and repeatable, just did it again.
15:00:51 <andrewsw> I'll see if I can attach gdb
15:01:35 <warlord> NEAT!
15:07:42 <andrewsw> I set the start and end dates to the same date (8/31/08) in a customized cashflow report.
15:08:22 <warlord> Hmmm...
15:08:28 <warlord> div-0 error?
15:08:55 <andrewsw> nope, well, maybe. it's related to the actual keystrokes.
15:09:58 <andrewsw> in the start date box, enter 8/31 <tab> <Tab> <Tab> should put you in the end date box, enter 8/31 <enter> <alt-O (the letter)> and it dies.
15:10:36 <andrewsw> in debian's 2.2.6
15:12:01 <warlord> What is Alt-O?
15:12:38 <andrewsw> short-cut for the okay button
15:13:27 <andrewsw> but if I click on ok, it works fine.
15:14:48 <andrewsw> works fine in relatively recent svn.
15:18:34 <warlord> Hmm
15:18:56 <andrewsw> huh. the problem lies specifically in the keyboard-shortcut.
15:19:30 <warlord> Hmmmm
15:19:56 <andrewsw> yup. whether the date is completed or not, using alt-O instead of clicking *while in the date entry box* causes a crash in both my custom or bog standard cashflow.
15:21:33 <andrewsw> does it in other reports too.
15:22:26 <andrewsw> gdb backtrace is mostly question marks. I'll paste one, but I don't think it's of any use.
15:25:47 <andrewsw> http://pastebin.ca/1195105
15:26:59 <warlord> Huh! Sitting in libguile!
15:27:42 <andrewsw> beyond me for sure...
15:29:07 <warlord> Do you have a libguile-debug package?
15:29:19 * andrewsw looks
15:34:16 <andrewsw> there doesn't appear to be a specific guile debug package, though I could be missing it, or it could be incorporated into some other package
15:41:54 <andrewsw> looks like I'd have to build a libguile-debug package from upstream... something I'm not going to do at the moment.
15:42:13 <andrewsw> I'll stick it in bugzilla with the backtrace.
15:45:55 <warlord> okay
15:51:31 <andrewsw> http://bugzilla.gnome.org/show_bug.cgi?id=551038 fwiw
16:00:32 *** sjc has joined #gnucash
16:04:54 <andrewsw> warlord: when you say libguile-debug, I assume you mean a version of libquile built with debugging symbols and *not* guile-debugging -> https://gna.org/projects/guile-debugging
16:05:12 <warlord> andrewsw: correct
16:39:54 *** PurityOfEssence has joined #gnucash
16:40:49 <PurityOfEssence> Hi is there an easy to way to associate income that is interest from my bank accounts with the accounts in gnucash
16:41:04 <PurityOfEssence> or need i create an interest account in income for each of my bank accounts?
16:42:03 <warlord> The latter
16:43:06 <andrewsw> okay, I've built what I think are debug packages for guile, but I get ERROR: permission denied "slibcat"
16:43:37 <andrewsw> do I need to reinstall slib over the top to get it properly registered?
16:44:26 <warlord> Run as root once.
16:44:33 <warlord> (to rebuild the slibcat)
16:44:47 <andrewsw> you mean run gnucash as root once?
16:45:06 <warlord> yes
16:45:40 <PurityOfEssence> if you run gnucash as root, do you get rich?
16:46:33 <andrewsw> all your cash are belong to us
16:47:01 <andrewsw> well, that's disappointing. no new info in the backtrace. I must not be doing something right.
16:47:02 <warlord> every time you run gnucash as root it paypal's $1 to me ;)
16:47:18 <andrewsw> oh... *that's* how we pay for support
16:47:23 <warlord> LOL
16:47:48 <andrewsw> building debug pckages for debian must be more complicated that a surface google suggests.
16:47:54 <andrewsw> I"ll have to mess with this later.
16:48:05 <warlord> with RPM it's easy.
16:48:26 <andrewsw> well, in theory, it's easy with debian too... but I must be missing something.
16:48:56 <andrewsw> I wonder though why I don't get the problem with svn
16:49:09 *** andi5 has quit IRC
16:49:36 <warlord> non-backported changes?
16:50:07 <andrewsw> well, sure, but I was sort of thinking that this problem was potentially lower-level than gnucash proper.
16:50:23 <andrewsw> and hence should show up in my build of svn as well.
16:50:53 <warlord> Could be?
16:52:37 *** PurityOfEssence has left #gnucash
16:53:37 <andrewsw> ah well, that's enough of that for now.
17:09:47 *** andrewsw is now known as andrewsw-afk
17:13:48 *** paulproteus_ has quit IRC
17:17:21 *** warlord is now known as warlord-afk
17:26:50 *** paulproteus_ has joined #gnucash
17:45:10 *** |gunni| has joined #gnucash
17:54:14 *** _gunni_ has quit IRC
18:38:16 *** cedayiv has joined #gnucash
18:44:55 *** warlord-afk is now known as warlord
19:01:06 *** martincleaver has quit IRC
19:18:50 *** esperegu has quit IRC
19:43:07 *** ErKa has joined #gnucash
20:28:03 *** ErKa has quit IRC
20:29:13 *** Zoolooc_ has quit IRC
20:37:21 *** cedayiv has quit IRC
21:19:49 *** donfede has quit IRC
21:30:00 *** sjc has quit IRC
22:18:21 *** ben_goodger has quit IRC
23:15:37 *** benoitg has joined #gnucash
23:47:58 *** warlord is now known as warlord-afk