2008-01-04 GnuCash IRC logs

00:18:21 *** w has quit IRC
00:25:59 *** twunder has quit IRC
00:55:40 *** tjanssen has quit IRC
00:57:58 *** andrewsw is now known as andrewsw-afk
01:03:27 *** XiXaQ has quit IRC
01:23:09 <d3ity> Hey is there anyone around that can walk me through setting up hbci for PNC bank?
01:23:23 <d3ity> i can't seem to figure out how to correctly add a user
02:36:03 *** XiXaQ has joined #GnuCash
02:39:27 <d3ity> Anyone? help with PNC OFX/HBCI?
02:39:46 <d3ity> not sure which option to select when creating a user
03:30:33 *** jjjjjjj has quit IRC
03:32:17 *** Zoolooc has joined #gnucash
04:00:56 *** Zoolooc_ has joined #gnucash
04:02:41 *** cj has joined #gnucash
04:02:44 <cj> hey folks
04:02:59 <cj> my mom (who is familiar with quickbooks) tried to use gnucash on windows:
04:03:00 <cj> http://colliertech.org/share/My_Gnu_Cash_Experience.html
04:04:52 <cj> warlord-afk: should I file a bug report?
04:04:54 *** kielein has joined #gnucash
04:07:49 <cj> http://bugzilla.gnome.org/show_bug.cgi?id=507209
04:09:55 *** Zoolooc has quit IRC
04:10:14 <cj> thanks for all your hard work. it looks like it's nearly there.
04:29:34 *** ErKa has joined #gnucash
04:48:45 *** ErKa has quit IRC
04:51:27 *** ErKa has joined #gnucash
05:21:46 *** cortilap has joined #gnucash
05:44:15 *** IanL has quit IRC
05:50:13 *** kling0n has quit IRC
05:52:48 *** kling0n has joined #gnucash
06:13:12 *** mcarter has joined #gnucash
06:21:42 *** Zoolooc__ has joined #gnucash
06:29:22 *** Zoolooc_ has quit IRC
06:29:55 *** Demitar has joined #gnucash
06:43:55 *** andi5 has joined #gnucash
06:43:56 *** gncbot sets mode: +o andi5
07:23:51 *** motin_0 has joined #gnucash
07:38:57 *** IanL has joined #gnucash
07:41:06 *** motin_0 has quit IRC
08:35:24 *** ErKa has quit IRC
08:45:17 *** wharp has joined #gnucash
09:13:55 *** Demitar has quit IRC
09:19:09 *** warlord-afk is now known as warlord
09:19:10 <warlord> .
09:19:10 <gncbot> warlord: Sent 9 hours and 50 minutes ago: <andrewsw> oh that's sneaky -- (lambda (l) (gnc:define-report...))
09:19:46 <warlord> cj: Looking at those experiences I think I see no bugs but a couple RFEs.. And I think we already have some of those in the system (please look)
09:19:49 <foo> Mornin' warlord
09:20:11 <warlord> d3ity: You need OFX, not HBCI, but beyond that I dont know. You'll need to find the OFX server information.
09:21:42 *** wizkid239 has joined #gnucash
09:27:02 *** donfede has joined #gnucash
09:29:30 *** wizkid238 has quit IRC
09:31:49 *** Demitar has joined #gnucash
09:38:23 *** ErKa has joined #gnucash
09:39:14 *** McLANE has joined #gnucash
09:39:51 *** tjanssen has joined #gnucash
10:12:51 *** twunder has joined #gnucash
10:34:00 *** Slart has quit IRC
10:41:26 *** billy__ has joined #gnucash
10:48:32 *** kling0n has quit IRC
10:50:13 *** kling0n has joined #gnucash
10:59:23 <billy__> Hello Folks, Is it possible to move all the transactions of one sub-account to another? For example, at the moment I have a sub-account called 'gas' within parent account called 'expense' I want to create a sub-account called 'fuel' within expense, and then move all the fuel relate sub-account of expense to this account. Hope this makes sense ...thanks
11:00:39 <cortilap> you could rename gas to fuel, then create a new gas expense and move fuel inside it
11:00:45 <cortilap> (if it's not possible to do a mass move)
11:01:02 <jsled> Or make Fuel the parent account of Gas (via Edit Account)
11:01:17 <jsled> If you then delete "Gas", it should prompt you to move all the transactions to Fuel.
11:01:28 <cortilap> neat
11:02:04 <billy__> cortilap and jsled : thanks... i'll give it a go :-)
11:04:04 *** IanL has left #gnucash
11:07:07 <billy__> jsled: worked a treat...thank you very much :-)
11:11:54 *** McLANE has quit IRC
11:20:23 *** andrewsw-afk is now known as andrewsw
11:20:56 *** wizkid239 is now known as wizkid238
11:23:20 <andrewsw> cj: re: #7 gnc doesn't have "memorized txns" in the way that quickbooks does. It has sched. txns, based on calendar events, and it recalls the most recent version of a particular txn. QB memorized txns are *like* the recalled ones only they are immutable. In gnc you only get the mutable, based on last entry, version.
11:23:26 *** andrewsw is now known as andrewsw-afk
11:32:35 *** kling0n has quit IRC
12:08:31 *** tobi has joined #gnucash
12:09:40 <tobi> hi, after the year changed to 2008 I am not able to open my account. If I switch the local date to 12.12.07 and open it, everything is working fine.
12:26:05 *** billy__ has quit IRC
12:35:23 *** fell has joined #gnucash
12:38:51 *** andrewsw-afk is now known as andrewsw
12:39:26 <andrewsw> tobi: in what way are you not able to open the account? does the app crash? please provide more details.
12:41:07 <tobi> andrewsw: uh. sorry. it segfaults.
12:41:56 <andrewsw> what version, what OS?
12:42:40 <tobi> gentoo, gnucash-2.2.2
12:43:08 <tobi> Do you want to have a strace?
12:44:12 <andrewsw> We'll want a copy of whatever console output there is, and possibly a gdb backtrace. use pastebin.
12:44:34 <andrewsw> I bet you've got a bad scheduled transaction.
12:45:06 <andrewsw> another option is to run gnucash using `gnucash --nofile` and then edit the preferences to turn off automatice creation of SX's. THen you can open the file and try to figure it out.
12:45:10 *** andi5 has quit IRC
12:46:32 <tobi> andrewsw: But why does it work to open the file when I set my date to 2007 and then reset the date to now?
12:46:40 <andrewsw> ?
12:46:54 <andrewsw> review the steps please.
12:47:31 *** kling0n has joined #gnucash
12:48:26 <tobi> after gnucash --nofile, editing preferences it lets me open my file.
12:48:48 *** Jaran has joined #gnucash
12:48:50 <jsled> do you have any scheduled transactions?
12:49:05 * andrewsw passes tobi to jsled and ducks
12:49:38 <tobi> yes. and after I have clicked on scheduled transaction -> since last run it crashed again
12:50:02 <jsled> Yeah. I don't think it's the date, so much as a bug that was new in 2.2.2
12:50:41 <andrewsw> I thought that was Mac specific and thus haven't been following.
12:50:42 <jsled> You could downgrade to 2.2.1, or I can come up with a patch + overlay ebuild for 2.2.2, if you like … ?
12:51:02 <jsled> No, it's not Mac specific.
12:51:05 <tobi> that would be great :)
12:53:13 * jsled works one up.
12:53:15 <tobi> can you explain in a few words what the problem is? I have only one scheduled transaction which is run every 1st of month...
12:53:38 <jsled> memory corruptino.
12:53:43 <jsled> corruption, even.
12:53:52 <andrewsw> oh, is *that* the GUID length?
12:53:56 <jsled> yes
12:54:12 <andrewsw> ding ding ding, /me gets it.
12:54:53 *** cortilap has quit IRC
12:54:59 *** _gunni_ has joined #gnucash
12:56:52 <jsled> tobi: do you already have a local overlay?
12:58:31 <tobi> sure.
12:58:34 <tobi> with layman
12:58:52 <jsled> but do you have a /usr/local/portage/ or something you can manipulate?
12:59:09 <tobi> yes, too
13:01:49 *** palatin has joined #gnucash
13:04:58 <jsled> Hmm. Damn patch does't want to apply.
13:07:00 <jsled> There we go.
13:08:24 <jsled> Right ... mkdir -p /usr/local/portage/app-office/gnucash/files
13:08:43 <jsled> cp /usr/portage/app-office/gnucash/gnucash-2.2.2.ebuild /usr/local/portage/app-office/gnucash/gnucash-2.2.2-r1.ebuild
13:09:23 <jsled> curl "http://pastebin.ca/raw/841483" > /usr/local/portage/app-office/gnucash/files/r16766.patch
13:09:51 <jsled> cp /usr/portage/app-office/gnucash/files/gnucash-2.2.2-icons.patch /usr/local/portage/app-office/gnucash/files
13:10:06 <tobi> done :)
13:10:07 <jsled> ebuild /usr/local/portage/app-office/gnucash/gnucash-2.2.2-r1.ebuild digest
13:10:15 <jsled> emerge gnucash
13:10:25 <jsled> That should do the trick.
13:10:30 <jsled> Oh, except.
13:10:31 <jsled> Heh.
13:10:55 <tobi> you have forgotten to include epatch r16...
13:11:21 <jsled> cd /usr/local/portage/app-office/gnucash; curl "http://pastebin.ca/raw/841486" | patch -p0
13:11:33 <jsled> (Or manually. whatever works for you)
13:12:33 <andrewsw> are we going to push out 2.2.3 quicker to deal with this problem?
13:12:47 <jsled> I've gotten that sense, yes.
13:13:19 <andrewsw> and is that separate from the apparently serious Mac issues? (or was that packaging, I can't keep track, yet)
13:13:40 <jsled> I'm not sure.
13:14:28 <andrewsw> is andi5 release manager now?
13:14:31 <tobi> jsled: I am now installing it. thank you for your effort!
13:14:33 <jsled> aye.
13:14:50 <jsled> tobi: you're welcome. Sorry it went out that way in the first place.
13:15:23 <tobi> no need to excuse. thank you for this great project!
13:21:19 <andrewsw> I can only assume the proliferation of QIF importer chatter is a result of gaining lots of new users...
13:26:05 *** dbreiser has joined #gnucash
13:26:50 <dbreiser> d3ity: see http://wiki.gnucash.org/wiki/Setting_up_OFXDirectConnect_in_GnuCash_2
13:27:26 <dbreiser> d3ity: click on the links ending in .png to see screenshots of the aqbanking setup wizard dialogs
13:28:45 <dbreiser> d3ity: see http://lists.gnucash.org/logs/2008/01/2008-01-02.html#T00:08:29 for the server address and a couple other ofx parameters
13:29:28 <dbreiser> d3ity: those parameters work as long as you are talking about PNC that started out as Pittsburg National
13:30:27 <dbreiser> d3ity: the PNC in Pennsylvania and surrounding states still uses the server address I posted. I double checked since the original post
13:30:54 <nbinont> andrewsw: I assume so - we were pushing 1000 dowloads/day for windows
13:31:29 <andrewsw> wow!
13:32:18 <dbreiser> d3ity: and they are using OFXDirectConnect. HCBI is almost exclusively German, but aqbanking is the source of both HCBI and OFXDC for gnucash
13:40:02 <dbreiser> andrewsw: It looks like it took the Linux users about a week longer than the mac users to run into the guid_length problem.
13:40:18 <dbreiser> I started getting reports less than a day after I posted 2.2.2
13:40:35 <andrewsw> hmm... interesting dynamic there, huh?
13:40:40 <dbreiser> yep
13:40:50 <dbreiser> I thought it was just me and Intel
13:41:17 <andrewsw> I'm not sure exactly when but it looks like deb sid only *just* got 2.2.2.
13:41:38 <andrewsw> maybe I should go file a pre-emptive bug over there.
13:41:42 <andrewsw> just to keep it out of lenny
13:42:17 <dbreiser> and what was weirder for me was that my machines only failed if gnucash was built from tarball -- no failure from svn
13:43:15 <dbreiser> but that might have been just the SX editor close crash, not the SLR crash
13:44:12 <dbreiser> gotta go
13:44:14 *** dbreiser has quit IRC
13:51:18 <andrewsw> should my report-guid work be committed directly to trunk or should there be a branch for it?
13:51:47 <andrewsw> I ask probably mostly out of nervousness and because it's still pretty *ugly* but appears to be mostly functional.
13:52:04 <andrewsw> well, frankly all the way functional in my testing.
13:52:33 <andrewsw> but it's not *done* with a ~700 line diff.
13:53:49 <fell> hi, andrewsw, I have to leave in short. Some question about my last devel posting?
13:54:55 <andrewsw> hi fell! I saw that, but I don't know the answer. That is, I don't know where that formatting happens.
13:55:25 <andrewsw> But if it's not formatted properly in the html report display, then we certainly can't expect it to work when exported.
13:56:04 <andrewsw> And even if it *does* get displayed properly, that doesn't mean the importing app will handle it properly.
13:56:32 <fell> No, the indians wish it simply unformated.
13:56:51 <andrewsw> So, yes, I think it should be an option to suppress super-fractional (made that up) numeric punctuation.
13:58:16 <andrewsw> I don't know how extensive those changes would have to be.
13:59:53 <andrewsw> just a quick glance at a report suggests it might be easy to do at the report level by changing the markup from "number cell" to some other format. maybe that would be sufficient to excise the comments.
14:00:07 <andrewsw> Maybe you could summarize our discussions in the relevant bug?
14:01:51 <fell> yes, but earliest after teh weekend. If you have time before, feel free to do it.
14:01:58 <andrewsw> okay.
14:02:04 <andrewsw> take care fell.
14:03:09 <fell> It doesn' look as if this weekend the sky collapses.
14:04:07 <fell> (from asterix, a belgian cartoon)
14:04:32 <andrewsw> haven't read that in forever...
14:05:28 <fell> It is really good. Have a nice weekend.
14:05:55 <andrewsw> you too.
14:06:14 *** fell has quit IRC
14:06:57 <andrewsw> jsled: do we have a good, reproducible example of the 2.2.2 SX crash? TB wants one before I file a release critical bug against 2.2.2
14:07:32 <jsled> Not really ... it's a memory corruption issue.
14:07:43 <jsled> So, it sometimes happens, sometimes doesn't.
14:11:24 <andrewsw> okay.
14:28:32 *** aluz has joined #gnucash
14:34:19 *** McLANE has joined #gnucash
14:35:18 <aluz> How can I add a new "Type" in the "Select Security Window" when dealing with stocks? My stocks are not in the NASDAQ, but in the BOVESPA. I'm with gnucash 2.0.5 (Debian stable)
14:35:42 <jsled> Just type it in.
14:37:28 <aluz> it's not possible to type.
14:38:36 <jsled> Ah, I think you need to create it in the Security Editor, first...
14:40:26 <aluz> Ok. There I can type. Many thanks.
14:45:50 <warlord> Yeah, you can type it after you click "New" ...
14:55:40 <aluz> "Get Quotes" doesn't work with BOVESPA or BVSP in place of NASDAQ. Anyone knows what can I use?
14:57:39 *** |gunni| has joined #gnucash
14:57:40 *** _gunni_ has quit IRC
14:59:29 <tobi> jsled: gnucash compiled fine. and everything is working now again. many thanks!
14:59:37 <jsled> aluz: which stock? Have you tried the "yahoo_brazil" quote source?
14:59:41 <jsled> tobi: great! :)
14:59:50 *** McLANE has quit IRC
15:00:01 <jsled> tobi: I'll file a gentoo bug ... hopefully a real 2.2.2-r1 will come down the pipe soon.
15:00:13 <jsled> frankly, I'm not quite sure how that'll interact with the fake one we just did.
15:00:29 <jsled> OF course, 2.2.3 will fix the problem for real, but that's going to take a few days, at least.
15:02:05 <aluz> jsled: I've tried yahoo_brazil with petr4 and others with no result.
15:03:24 <jsled> "yahoo_brasil", I mean. And it seems that "petr4" works for me.
15:08:34 <aluz> BOVESPA:petr4 does not work for me. I'm using the "stock editor" and the "security editor" and trying a lot of possibilities.
15:09:58 <aluz> jsled: do we have the same version - 2.0.5?
15:10:18 <jsled> Actually, I'm just using gnc-fq-dump...
15:10:46 <jsled> http://pastebin.ca/841627
15:12:15 <jsled> What are you entering as the Symbol/abbreviation?
15:12:49 <aluz> gnc-fq-dump works for me too.
15:13:14 <aluz> symbol - petr4
15:13:43 <jsled> Hmm. Of course, the date is all messed up.
15:14:08 <jsled> (in the gnc-fq-dump output)
15:15:24 <jsled> Yeah ... `echo -ne "(yahoo_brasil \"petr4\")" | gnc-fq-helper` returns "(#f)"
15:15:27 <aluz> I just tried PETR4.SA as Symbol/abreviation with no success.
15:15:35 <jsled> Sorry. The yahoo_brasil parser appears broken. :(
15:16:54 <aluz> ok. It's not the first software I try. Many thanks anyway.
15:19:43 <andrewsw> I didn't know we still load saved-reports-1.8
15:20:09 <cj> andrewsw: I'll tell mom
15:20:28 <andrewsw> cj: does that make sense to you?
15:22:22 <cj> thanks warlord. I'll do a lookup when I get a chance.
15:22:32 <warlord> ok.
15:22:41 <cj> andrewsw: it does to me. hopefully it makes sense to my mom, too :)
15:22:44 *** Zoolooc__ has quit IRC
15:23:12 <cj> I'll point her at the GNOME bugzilla and help her start entering RFEs and bugs herself
15:23:26 <andrewsw> memorized txns is a feature I've wanted for a while as a way to overlay the automatically created list of auto-fill txns
15:23:55 <andrewsw> that would allow the user to specify exactly how they want txns to payee X look as a default even though once in a while it might be different.
15:24:21 <andrewsw> The way it currently exists, if the last one entered was the odd-ball version, you have to correct it to the preferred version.
15:24:58 <jsled> I hit that all the time, with Groceries, where I basically have 2 templates: weekly shopping vs. mid-week.
15:25:28 <jsled> Which have very differen splits.
15:25:45 <andrewsw> I'd love to memorize the preferred version and then after the autofill list gets populated, replay the memorized list over the top of it.
15:25:57 <jsled> I'd rather that it always populated the superset of relevant splits, and just silently removed the zero-credit/debit ones at the end.
15:26:12 *** aluz has left #gnucash
15:26:41 <andrewsw> hm... I like the zero-credit/debit ones for connecting dividends -> cash back to the relevant stock/fund acct...
15:27:18 <andrewsw> see svn r16774
15:27:46 <jsled> Maybe not zero, then ... blank is better.
15:28:42 <andrewsw> oh, does it differentiate? I was actually using blanks, but it doesn't matter to me as long as we get to keep that split that ties back to the stock acct.
15:28:54 *** tobi has quit IRC
15:39:31 *** warlord is now known as warlord-afk
15:48:09 *** wizkid238 has quit IRC
15:49:17 *** wizkid238 has joined #gnucash
16:19:06 *** wharp has quit IRC
16:24:38 *** warlord-afk is now known as warlord
16:29:35 *** kling0n has quit IRC
16:43:58 <andrewsw> wOOt!
16:45:01 *** Demitar has quit IRC
16:46:15 <andrewsw> I have three options for warning about an old-style saved report:
16:46:27 <andrewsw> 1) put a gnc:warn in for console output.
16:46:52 <andrewsw> 2) pop up a gnc-error-dialog for each report (cumbersome for large numbers of reports, or
16:47:14 <andrewsw> 3) find a way to get the info back out and warn once that there are "some" old reports.
16:47:24 <andrewsw> 1 is easy to code.
16:47:34 <andrewsw> but would be missed by most users
16:48:04 <andrewsw> 2) is easy to code, but is a royal PITA to users with large numbers of reports that may be perfectly fine for a long time.
16:48:34 <andrewsw> 3) is a royal PITA to code or entails yet another messy data field
16:48:56 <andrewsw> but provides a clean, one click warning for the user, with high visibility, but low level of interference.
16:49:21 <warlord> why is 3 so hard to code? just have a global boolean for whether or not you've raised the dialog.
16:49:31 <warlord> If you're good, reset the boolean as part of the book close hook
16:49:57 <andrewsw> hmmm.. beacuse I'm an idiot?
16:50:28 <warlord> Heh
16:50:30 <andrewsw> Where would one put such a boolean? report-system.scm?
16:51:00 <andrewsw> ...must... make... ideas... fit ...in ...box...
16:52:44 <warlord> sure. why not?
16:52:46 <andrewsw> I mean, I can't just wrap a (let) around the whole shebang. Can I (define old-reports #f) (export old-reports) and then later (set! old-reports #t)?
16:53:02 <andrewsw> is a (define) mutable with (set!)?
16:54:19 <andrewsw> no need to reset on book close because the saved reports are loaded once per instance of gnucash.
16:55:29 <andrewsw> google answers.
16:55:33 *** Demitar has joined #gnucash
16:56:00 * andrewsw 's weaknesses show...
17:01:28 <warlord> Yes
17:02:43 <andrewsw> what I can't find is the entry point for the report system. I want to put that warning at the end of the process of loading the reports...
17:04:42 *** tjanssen has quit IRC
17:06:23 <warlord> Doesn't matter where you put it.
17:06:40 <warlord> The bool will make sure you only pop up the dialog once.
17:06:58 <andrewsw> yes, but...
17:07:02 <andrewsw> oh. damn that box.
17:07:05 *** andi5 has joined #gnucash
17:07:05 *** gncbot sets mode: +o andi5
17:07:18 <andrewsw> We only need to warn them once and it doesn't matter which report it keys on, just that it warns.
17:07:25 * andrewsw crawls under a rock
17:07:46 <warlord> :-D
17:15:51 *** twunder has quit IRC
17:16:50 <andi5> andrewsw: is there anything you wanted to ask me about? :-) (just skimmed the backlog)
17:17:02 <warlord> I think he was wondering about 2.2.3
17:17:11 <andi5> nbinont: ping
17:17:32 <andrewsw> I was curious what I could do to help push out 2.2.3 since 2.2.2 is kind of problem.
17:18:15 <andi5> hm... make sure you have everything up that should be included :-)
17:18:36 <andi5> oh, and you could audit for me: http://lists.gnucash.org/logs/2008/01/2008-01-03.html#T17:59:36
17:20:17 <nbinont> hi andi5
17:20:26 <andi5> nbinont: nice, very fast :-)
17:20:33 <nbinont> got lucky
17:20:55 <andi5> nbinont: what is your time frame this sunday or at the beginning of the coming week?
17:21:04 <andi5> maybe not sunday
17:21:48 <andi5> i honestly ask you, that is in no way a demand or something :-)
17:22:11 <andi5> someday i will try to ask earlier again :)
17:22:22 <nbinont> Just trying to see how I'll fit in
17:22:25 <andi5> and maybe i should do it on gnucash-devel again
17:22:43 <andrewsw> andi5: which ones, the snv revisions or the gcc 4.2 fixes?
17:22:48 <andrewsw> svn
17:23:04 <nbinont> andi5: this weekend isn't great - school starts monday
17:23:28 <andi5> i think the first 4.2 fix is included in the list, ....but feel free to give us some input on the linked thread as well :)
17:23:30 <nbinont> but I should know better tomorrow
17:23:42 <andi5> nbinont: ok
17:24:11 <andi5> nbinont: i guess i will downgrade gnome-vfs...
17:25:38 <nbinont> andi5: I'll have time sunday (but not sure about the internet connection), monday night can work too
17:26:10 <andi5> nbinont: what time utc would you need the tarball on monday?
17:27:58 <nbinont> andi5: preferably by 13:00 utc
17:28:36 <andi5> ok... let us defer the decision until tomorrow, ok?
17:28:45 <nbinont> andi5: ok
17:30:15 <andi5> ot, why does thomas ask questions about aqbanking? i supposed he was a debian evangelist :-)
17:30:57 <andrewsw> heh heh.
17:31:44 <andrewsw> be nice, I convinced him to patch 2.2.2 in sid, before it hits lenny, so we won't have months of debian fanatics complaining about SX's crashing...
17:32:14 <andi5> 2.2.3 is unrealistic?
17:33:17 <andrewsw> if 2.2.2 slips into lenny and 2.2.3 hits sid, 2.2.3 will sit in sid for at least 10 days (assuming no serious bug reports against it) before it moves into lenny.
17:33:24 <nbinont> andi5: come to think of if...I'd probably only know better by 22:00 utc tomorrow, so that doesn't help you much. How about this - If you decide on 2.2.3 this weekend, I'll make the windows build happen monday.
17:34:08 <andrewsw> if 2.2.3 gets significant bug reports, it could sit in sid for a *long* time leaving a whole mess of lenny users with a crasher bug.
17:34:25 <andi5> i see
17:34:47 <andi5> nbinont: seems like you just decided you may saturday will like alike ;-)
17:35:14 <andi5> s,[^a]like, look,
17:35:30 <andi5> no, that does not work ;-)
17:35:39 <andrewsw> LOL
17:36:18 <andrewsw> Colorless green ideas sleep furiously.
17:37:15 <andrewsw> 16776 looks a-ok, though you have to wonder why it was done the way it was... is some other code relying on it being formatted in the old stupid way?
17:37:50 <nbinont> andi5: lol, yah - just thought up a backup plan for the internet.
17:39:25 <andi5> andrewsw: always a good question... but i really hope this is not true... as long as the reverse function is a real inverse :)
17:51:01 <andrewsw> it looks like the inverse was broken by the old +04-30 method. My C is bad, but it looks like it would have just ignored the minutes if it hit the '-'.
17:52:38 <andrewsw> I don't know if that was ever used that way, but, that's how it looks.
17:57:51 <andi5> yep, that is what i think as well... and i think it should work perfectly now
17:59:24 *** palatin has quit IRC
17:59:38 <andrewsw> 16771. I remember hitting this after submitting some earlier changes for gcc-4.2 and poured over the code trying to figure out how the heck to fix it.
18:00:14 <andi5> one stupid variable did the trick :-)
18:00:15 <andrewsw> gcc-4.2 didn't like that the const parms were getting passed on without the const. I followed it downstream a long way and it just kept getting uglier.
18:00:19 <andrewsw> yup.
18:00:45 *** nbinont is now known as nbinont-afk
18:00:51 <andrewsw> I think it's good too, though I don't know that it's BP worthy. But that's because I don't really know what merits BP.
18:01:19 <andrewsw> I guess it could cause packaging systems to fail, but do packaging systems build with warnings on?
18:01:43 <andi5> i am not yet sure when 2.4 will hit the surface, so i suppose 2.2 should support new compilers
18:02:21 <andrewsw> sure. Who are we trying to support with new compilers though?
18:02:36 <andrewsw> If its distro packagers, I can't imagine they build everything with warnings on.
18:02:51 <andrewsw> If its users building, then its probably a good idea.
18:03:07 <andrewsw> is "less noise for dev's" a valid BP justification? ;)
18:03:13 <andi5> i suppose the latter
18:03:15 <andi5> i think so :-)
18:03:37 <andi5> the same reason for the goffice-0.5 backport... but hey, did anyone vouch for that besides me? ;-)
18:04:11 *** wizkid239 has joined #gnucash
18:04:20 *** wizkid239 has left #gnucash
18:05:12 <andrewsw> 16755 is an obvious data-loss preventer.
18:06:01 <andrewsw> looks like it completes work that was started before in terms of stripping invalid characters.
18:06:20 <andrewsw> But, out of curiousity, why strip them instead of escape them?
18:07:03 <andi5> i suppose because that was easier to implement -> edit the string in-place ;-)
18:08:05 <andi5> actually, i do not know
18:08:26 <andrewsw> ISTM, assuming stripping non-utf8 is the right move, then any source of data should be stripped, and these couple of fields weren't being stripped, so stripping them is the right move.
18:08:50 <andrewsw> But the greater question is, are we losing data by stripping instead of escaping those characters?
18:09:00 <andrewsw> But I don't really understand character encoding issues.
18:09:33 <andrewsw> I have no comment on 16742.
18:09:42 <andi5> sure, we lose the raw bytes....
18:10:07 <andrewsw> the comment says "Validate the input strings to ensure utf8"
18:10:20 <andrewsw> if we are ensuring utf8, then we can't have non-utf8 chars.
18:10:46 <andrewsw> by that rationale, they should be stripped.
18:10:50 <andi5> if the job is to validate, then we should return a boolean :-)
18:10:58 <andrewsw> ha ha!
18:11:13 <andrewsw> En-validate?
18:12:13 <andi5> why that? ... \xfe as 4-character, 4-byte utf-8 string is utf-8 as well
18:12:59 <andrewsw> I was going to ask that, is an escaped non-utf8 character considered a valid utf8 char, then hmmm..
18:13:10 <andrewsw> s/is/if/
18:13:51 *** kielein has quit IRC
18:15:16 <andi5> that is just yet another possible enhencement bug... throw an error, fetch it, present the user a dialog to pick an encoding, strip or escape...well... but hey
18:15:26 <andrewsw> faced with losing complete txns versus losing a handful of bytes, the choice is obvious.
18:15:37 <andi5> that is for sure
18:15:57 <andrewsw> My concern is what happens when someone gets a wacky file with refnums that are *all* non-utf8... don't know if such a thing could happen.
18:17:04 <andi5> well, i highly doubt the example byte of the bug was relevant.... so .... it probably does not hurt
18:17:34 <andi5> i think, it was something superpositioned in N>o<. xyz
18:18:10 <andi5> superior, or whatevre
18:18:37 <andrewsw> superscript.
18:18:44 <andrewsw> I agree.
18:19:15 <andrewsw> So I think it's good.
18:19:19 <andrewsw> and should be BP'd
18:19:29 <andi5> andrewsw: how many guys do you know that use encodings differing a lot from utf-8 but are not sensitive to encodings?
18:19:42 <andi5> they probably know how to convert files :)
18:19:50 <andrewsw> yes.
18:20:16 <andi5> (after or before importing in gnucash)
18:20:31 <andrewsw> most folks don't know jack about encoding. I know I sure don't
18:20:33 <andi5> ah well, just ignore me
18:20:46 <andi5> i mean... when i expect problems, i write Koehler
18:20:48 <andrewsw> so the ones that know should know how to handle it.
18:21:05 <andrewsw> ö
18:21:23 * andrewsw loves middle-click paste
18:21:26 *** localhost has quit IRC
18:21:27 <andi5> hehe
18:21:49 *** localhost has joined #gnucash
18:23:38 <andrewsw> as to the other compiler fixes. I agree with the later comments that it's better to *not* hide the warnings. As to the OP's proposed fixes, I haven't a clue.
18:23:54 <andrewsw> "As to" WTF.
18:24:28 <andi5> wtf is wtf? :-)
18:24:42 <andrewsw> LOL
18:25:04 * andi5 remembers http://bash.org again
18:27:47 <andrewsw> http://bash.org/?search=andi5&sort=0&show=25, no results returned.
18:28:40 <andrewsw> http://bash.org/?search=warlord&sort=0&show=25 returns http://bash.org/?8902
18:29:14 * andrewsw has a new way to help the release manager... distraction!
18:29:49 <warlord> Not me.
18:29:54 <andi5> at least i won some happiest irc channel user of the month, typing the most smilies :-D
18:30:14 <andrewsw> warlord: yeah, I figured.
18:30:20 <foo> http://slashdot.org/ = borked, quick, look now!
18:30:49 <andrewsw> ??
18:30:52 <warlord> B0rked how?
18:30:54 <andi5> looks like always :-)
18:31:07 <andi5> messed up, pretty usual
18:31:08 <warlord> Other than that I can't read it because the BG and text are both dark
18:31:17 <andrewsw> mm-bork-bork mm-bork-bork
18:31:19 <jsled> who ever goes to websites anymore? :)
18:31:26 <andrewsw> that's so web2.0
18:31:27 <andi5> jsled: me
18:31:36 <andi5> LOL
18:31:56 <andi5> andrewsw: you mean semantic web is stone age?
18:33:05 <andrewsw> It's a perfect storm of actualized paradigm shifts in a ... /me peters out...
18:33:53 <warlord> BINGO@!
18:36:08 <andrewsw> I want to get some more eyes on this report-guid. Should I push it up to trunk, or would you prefer I use a branch?
18:36:27 <jsled> just commit it. It can be backed out/fixed if need be.
18:36:31 <warlord> It's trunk... If it's working for you, commit it.
18:36:40 <andrewsw> wee....
18:36:44 <andi5> make it compile and everything is fine
18:36:58 <warlord> Well, it's scheme.. there are no compile checks.
18:37:06 <andi5> yeah, so the job is done :)
18:37:13 <andrewsw> hehehehe
18:37:17 <warlord> Would be nice if we had a "make check" that could at least load the reports to make sure there were no syntax errors.
18:37:54 <andrewsw> ha! not in *my* code...
18:38:25 <foo> hehe
18:38:30 <andi5> i suppose one could even match output for input example files & options against some given file contents?
18:40:21 <andrewsw> that assumes the test output is correct.
18:40:25 <warlord> Well, it would be nice to be able to actually RUN the reports and make sure they output reasonable data, but I'm less concerned about that.
18:41:30 <andrewsw> Seems a lot of the report work involves changing the output of the reports, so I don't know how far that would go.
18:41:33 <andi5> warlord: i am not into reports... what do you mean by "load the reports"?
18:42:27 <warlord> andrewsw: yeah, I think testing the reports actually RUN is a much harder problem.. Maybe just a search that the output ISNT "Report failed".
18:42:38 <andrewsw> just what I was thinking.
18:42:42 <warlord> andi5: which word didn't you understand?
18:42:49 <andrewsw> "the"?
18:43:01 <andi5> load the reports module? i suppose no :)
18:45:54 <andi5> well, it is quite unlikely i will add those test, so...
18:50:52 <warlord> Load each report .scm file
18:51:14 <andi5> oook
18:54:52 <andi5> andrewsw: thanks a lot for your help!
18:54:59 <warlord> Nice work, andrewsw
18:55:03 <andrewsw> ?
18:55:08 <andrewsw> you guys are too nice.
18:55:12 <andi5> hehe :)
18:55:20 <andrewsw> I need to go over to #wmii and get harrassed.
18:55:50 <andrewsw> they're downright mean over there.
18:56:08 <nbinont-afk> #wmii?
18:56:12 *** nbinont-afk is now known as nbinont
18:56:29 <andrewsw> a tiling window manager.
18:56:50 <nbinont> ah
18:57:21 <warlord> I'm waiting for a response to a question in #svk over on freenode.
18:57:39 <andrewsw> how long you been waiting?
18:58:06 <warlord> 30 minutes
18:58:14 <andrewsw> heh.
18:58:20 <andi5> warlord: try to ask anything on #git... they are quite fit and reponsive
18:58:33 <warlord> But that wont answer my svk question ;)
18:58:48 <andi5> something along the lines "use git instead" ;-)
18:58:53 <andrewsw> sure they will: rpm -i git
19:00:33 *** nbinont is now known as nbinont-afk
19:01:26 * andrewsw makes a note to give warlord $1MM
19:02:41 * foo intercepts money
19:03:28 <andi5> ... and wastes it at the bar
19:03:51 * andrewsw counts the profits gained from foo spending money at his bar.
19:03:58 <foo> haha
19:04:13 <andi5> andrewsw: positive?
19:04:30 <andrewsw> of course! How do you think I have time to hack gnucash?
19:04:55 <andi5> hm.... you got point
19:05:39 <andrewsw> not enough to give warlord $1MM though. :(
19:08:58 *** donfede has quit IRC
19:17:32 *** ErKa has quit IRC
19:18:24 <warlord> heheh
19:22:33 *** nbinont-afk is now known as nbinont
19:23:13 <nbinont> andi5: you don't happen to use google calendar?
19:23:24 <andi5> nope... should i?
19:23:48 <nbinont> andi5: not unless you want to ;)
19:23:56 <andi5> ok :-)
19:27:17 *** wharp has joined #gnucash
19:28:07 *** nbinont is now known as nbinont-afk
19:31:53 <andrewsw> hmmm... report-guid will not be backwards compatible. Users will not be able to open reports with older versions. How much of a problem is this?
19:34:48 <andrewsw> In fact, it will be a crasher bug if there are open reports in a report-guid supporting books file and it is opened in an older version of gnucash.
19:35:01 <andrewsw> I suppose this is mostly a problem for those of us running multiple versions...
19:36:09 <warlord> Is there a way to make 2.2 ignore it without crashing?
19:37:06 <andrewsw> yeah, probably. I could patch 2.2 to ignore it. Should be straightforward -- add the new fields to the record-constructor and that should do it, I think.
19:38:52 <andrewsw> hmmm... nope. the currently open reports would still fail disastrously.
19:39:18 <andrewsw> the saved reports would ignore the field, no problem. it's the open ones.
19:42:11 <andrewsw> yeah. its a crasher opening report-guid books file in 2.2.1
19:42:36 <andrewsw> Unbound variable: gnc:Report-template-new-options/report-guid
19:42:42 <andrewsw> a new function added as part of my changes.
19:44:28 <andrewsw> I had provided parallel functions for reading new and old style functions and use the new functions for storing new style functions.
19:44:45 <andrewsw> reports.
19:48:24 <warlord> well, what if you add gnc:Report-template-new-options/report-guid to 2.2?
19:48:32 <warlord> (that doesnothing)
19:49:06 <andrewsw> it can't do *nothing* because that is what returns the options to the report. Without it, the whole app crashes.
19:49:24 <andrewsw> Obviously that's another concern that should be addressed.
19:49:31 <warlord> Oh. Huh.
19:49:36 <warlord> If can't return #f?
19:50:05 <andrewsw> nope. The existing report-template-new-options/name does that if it can't find the template and it brings the whole app down.
19:50:31 <andrewsw> That's the cause of: http://bugzilla.gnome.org/show_bug.cgi?id=505921, essentially.
19:51:00 <andrewsw> So I guess the thing to do is fix 2.2 so that a #f isn't a crasher and provide the dummy function.
19:51:31 <andrewsw> but I have to run a kid to b-ball practice. So we'll have to discuss it more later.
19:51:35 <warlord> hmm.. what if it returns '() instead?
19:51:51 <andrewsw> ooh... that might work. good idea, you crazy box escaper!
19:52:08 <andrewsw> I'll chew on that while I watch her dribble around...
19:52:09 <andrewsw> later
19:52:11 *** andrewsw is now known as andrewsw-afk
20:12:53 *** wharp has quit IRC
20:17:27 *** wharp has joined #gnucash
20:18:51 *** IanL has joined #gnucash
20:28:23 * andi5 continues tomorrow
20:28:30 *** andi5 has quit IRC
21:00:25 *** |gunni| has quit IRC
21:13:13 *** BinaJr has joined #gnucash
21:28:33 <BinaJr> The Santander Banespa Bank (From Brazil) is generating the .ofx file with some trasactions sharing the same <FITID> <CHECKNUM> fields.
21:28:33 <BinaJr> If the transactions are inside the same .ofx file, everything ok.
21:28:33 <BinaJr> But if the trasactions are in separate files (for example, "the next month" file), gnucash completly ignores the existence of the transactions, even if they have different date and ammount values.
21:28:33 <BinaJr> I'm using gnucash 2.2.2 for Windows.
21:28:34 <BinaJr> Im trying to contact the bank to fix this. But, shouldn't gnucash to display any kind of alert message? Or to give the option to import the transaction?
21:29:31 <warlord> No, the bank is violating the OFX specification
21:29:43 <warlord> FITID is supposed to be unique per bank account.
21:31:23 <BinaJr> Yes, I know. The bank is wrong in generating. I'll give you an example. There was a kind of tax here (CPMF). Every transaction related to the CPMF has FITID equal zeroes.
21:32:20 <BinaJr> But, if the transaction has same FITID, but different values and different dates, shouldn't gnucash warn us about it? Should it offer the option of importing the """corrupted""" transaction?
21:33:15 <warlord> Arguable.
21:33:24 <lindi-> warning would be nice
21:34:40 <warlord> Send a patch
21:36:25 <lindi-> debian unstable issues prevent me right now :)
21:36:39 <warlord> well then. ;)
21:36:42 <BinaJr> I have not done any code for gnucash yet... I'll try to figure out my through the code and send a patch.
21:37:26 <BinaJr> ... out my through = figure out my way through the code ...
21:38:06 <warlord> src/import-export/ofx would be the place to look.
21:38:18 <warlord> you can add a call to gnc_error_dialog()
21:38:27 <warlord> (or ... whatever)
21:38:55 <BinaJr> Thanx! It saved me a lot of time.
21:39:01 <warlord> :-)
21:39:41 <lindi-> just managed to boot my laptop with a custom battery pack, was a real challenge to get it fit inside. (no i don't have the original pack so i can't just replace cells)
21:47:08 *** ttick has quit IRC
22:21:07 *** andrewsw-afk is now known as andrewsw
22:24:31 *** twunder has joined #gnucash
22:40:28 <warlord> andrewsw: FYI, I'm about to edit a report.
22:40:52 <andrewsw> which one, in what what?
22:40:53 <andrewsw> way?
22:41:01 <andrewsw> and hi!
22:41:15 <warlord> income-statement.scm
22:41:28 <warlord> Basically, re-adding "Profit & Loss" as an alternate name for the report.
22:41:50 <andrewsw> you know, that's fine. My only change is adding a 'report-guid to the define-report call.
22:42:23 <andrewsw> I have to roll back and fix my backwards compatibility anyway.
22:42:39 <warlord> Well, you'll have to add that to the new define-report call I'm adding.
22:42:58 <andrewsw> I want to make one BP ready commit before the big one.
22:43:04 <warlord> Okay.
22:43:06 <andrewsw> Okay, no problem.
22:43:39 <andrewsw> thanks for the warning!
22:45:53 <warlord> You're welcome.
22:48:18 <warlord> r16803.
22:48:24 <andrewsw> okay.
22:48:40 <warlord> I was tired of people not looking at Income Statement.. I think they well understood Profit & Loss.
22:49:13 <andrewsw> yeah. Of course going forward, it'll be easy to just rename the report without breaking existing open/save ones...
22:49:31 <warlord> True, but might as well allow both names.
22:50:19 <andrewsw> yep. but adding another define-report actually makes it two distinct reports. A saved report made with P&L won't work over Income Statement, because they have different GUID's
22:50:22 <warlord> (they need to be named differently for different things ... )
22:50:36 <warlord> This is true.
22:50:43 <warlord> But does it matter?
22:50:52 <andrewsw> don't know. probably not.
22:53:29 <andrewsw> hmmm, '() doesn't work as a return for the options.
22:53:36 <warlord> I mean, there are four define-reports for the various bar/pie charts..
22:53:39 <andrewsw> I need to return an empty set of options.
22:53:41 <warlord> No? Darn
22:54:11 <andrewsw> define-reports: yeah, it's just an interesting side-note, I guess.
22:54:50 <warlord> Yeah.
22:54:54 <andrewsw> options: maybe I can just call gnc:options-new to get a new options database
22:55:07 <andrewsw> make that gnc:new-options
22:56:04 <warlord> maybe....
22:56:30 <andrewsw> I need something that will *act* like real options.
22:56:46 <andrewsw> or, I could just write the name back out along with the guid.
22:56:53 <andrewsw> duh.
22:57:49 <warlord> Hmm...
22:58:04 <andrewsw> no.
22:58:13 <andrewsw> maybe
22:58:52 <warlord> hehe
22:58:57 <warlord> I leave it up to you...
22:59:17 <andrewsw> yes. I can write out both the guid and the name, tweak the .../report-guid functions to take two parms, throw away the one it doesn't use and call .../name
22:59:21 <andrewsw> thinking aloud.
23:01:59 <warlord> okay...
23:02:20 *** BinaJr has quit IRC
23:03:40 *** needo has joined #gnucash
23:04:56 <needo> I just started playing with gnucash today and really like it. I am switching over from Quicken and I was wondering if there is a report I can run that will provide me the same information as Quicken does from the "Account Balance Graph"
23:04:58 <andrewsw> just once(!) I'd like to remember to export a function first
23:05:21 <andrewsw> needo: what does that graph show?
23:05:43 <needo> It shows how much money you have at the end of the month left over after all your bills are paid.
23:06:44 <warlord> Well, there's an Asset Barchart.. Or the Income Expense Chart.
23:07:06 <andrewsw> It sounds like that report looks forward, does it include unpaid bills?
23:07:21 <warlord> Anyways, bedtime for me.
23:07:23 *** warlord is now known as warlord-afk
23:07:24 <needo> andrewsw: Yes.
23:07:32 <andrewsw> cya warlord
23:07:40 <andrewsw> andrewsw: then no.
23:07:47 <andrewsw> um. needo, that was for you.
23:07:53 <needo> hehe
23:08:09 <needo> well that sucks. that was a great planning tool.
23:08:40 <andrewsw> it's a much requested feature but is not likely to happen any time soon.
23:09:00 <andrewsw> there are other ways to get that information though --
23:09:42 <andrewsw> if you are using the vendors system, with invoices etc, you can report on how much is owed, for example. And because it tracks those items in a liability account,
23:10:19 <andrewsw> you could customize a balance sheet report to show your checking/savings accts and that one liability acct. That would give you the same sort of snapshot.
23:23:25 *** tjanssen has joined #gnucash
23:38:05 *** ttick has joined #gnucash
23:53:37 <d3ity> Can someone help me out with PNC and online banking? I can't seem to get my user settings right
23:54:40 <d3ity> https://www.oasis.cfree.com/test.ofxgp is what is listed as the server... however I'm unsure whether to create a keyfile or use PIN/TAN
23:55:10 <d3ity> i've tried both with various settings and nothing has worked... does anyone currently have PNC bank working properly with gnucash?
23:57:18 <andrewsw> d3ity: your timing isn't good. most of the devs are CET or US time, so you're probably too late. You might try posting to gnucash-user.
23:57:44 <andrewsw> @tell andi5 I see how it is... blame me for those BP's eh?
23:57:44 <gncbot> andrewsw: The operation succeeded.
23:58:23 <d3ity> andrewsw, my work schedule is slightly crazy... I'm US EST right now
23:59:01 <andrewsw> heh. sorry to hear that...
23:59:29 <d3ity> just got done with an 11 hour shift... and I'd like my computer to count my money for me...lol
23:59:40 <andrewsw> what bank is PNC?