2009-02-02 GnuCash IRC logs
00:02:48 <ska> anyone awake?
00:03:22 <ska> How can I change the default stylesheet for invoices? Can that be done? Why can't i set it to Fancy permanently??
02:05:28 *** Zoolooc has quit IRC
02:34:15 *** bentob0x has joined #gnucash
03:08:20 *** MechtiIde has joined #gnucash
03:25:49 *** ErKa has joined #gnucash
05:22:27 *** Woow has joined #gnucash
05:23:36 *** Woow has quit IRC
07:01:20 *** ErKa has quit IRC
07:17:45 *** twunder has joined #gnucash
07:27:42 *** mwic has quit IRC
07:31:14 *** IanL has joined #gnucash
07:31:20 *** Jimraehl has left #gnucash
07:31:49 *** JimRaehl2 has joined #gnucash
07:33:31 *** JimRaehl2 has left #gnucash
07:33:41 *** edmont has joined #gnucash
07:33:47 <edmont> hi
07:34:31 <edmont> i dont know how to do something
07:34:46 <edmont> i had a negativa value in a credit card account
07:35:04 <edmont> that is, the credit card owns money to me
07:35:45 <edmont> now, on february 1st, that value passed to the ordinary bank account
07:35:53 *** twunder has quit IRC
07:36:24 <edmont> then, value increases in the credit card accoun and ALSO increases in the ordinary account
07:36:32 <edmont> how can i solve that?
07:37:39 *** ErKa has joined #gnucash
07:40:20 *** edmont has quit IRC
07:44:45 *** JimRaehl2 has joined #gnucash
08:13:51 *** warlord-afk has quit IRC
08:14:11 *** warlord has joined #gnucash
08:19:16 *** nomeata has joined #gnucash
08:20:24 <warlord> ska: there is no way to change which report gets called when you click the 'print' button.
09:36:14 *** ErKa has quit IRC
09:42:10 *** warlord2 has joined #gnucash
09:46:20 *** warlord2 has quit IRC
09:46:52 *** warlord2 has joined #gnucash
09:47:28 <warlord2> andrewsw, jsled: could one of you op gncbot, please?
09:49:20 *** warlord has quit IRC
09:50:37 *** warlord2 is now known as warlord
09:51:42 *** jsled sets mode: +o gncbot
09:51:46 *** jsled sets mode: +o warlord
10:00:02 <warlord> thanks jsled
10:00:11 <warlord> hey, did you see my SX question from yesterday?
10:05:22 <jsled> yup
10:05:26 * jsled looks
10:08:07 <jsled> warlord: yup, doesn't work for me either.
10:08:19 <warlord> ah, good. so it's not just me.
10:10:50 * warlord is testing trunk to see if test-lots fails there, too.
10:19:10 <warlord> Ah, and that test is running now... and.... survey says....
10:20:16 <warlord> pass. so the failure is only in 2.2. Hmm...
10:22:46 <joslwah> By trunk you mean svn?
10:29:28 <warlord> it's all svn
10:29:34 <warlord> trunk branch, vs. 2.2 branch
10:30:18 <warlord> So test-lots only fails in the 2.2 branch. It passes in trunk...
10:30:33 <warlord> Which means time to binary-search and figure out when it failed.
10:33:56 <joslwah> Have fun.
10:34:18 <warlord> Oh yeah..
10:34:20 <warlord> So much fun.
10:34:31 * warlord is hoping jsled can look into the SX issue for me. ;)
10:34:55 <jsled> I'm probably not going to.
10:35:19 <warlord> :-(
10:35:28 <warlord> Too busy?
10:39:31 <joslwah> I'm vaguely thinking about rewriting the whole of the SX stuff, trying to clean it up, and add the ability to have dates +/- n business days.
10:40:00 <joslwah> Then we'd be able to more easily do things like predicated balance in n-days/weeks/months time.
10:40:17 <joslwah> But, I'm still in the contemplating stage, and not sure how far I'll get.
10:40:19 <warlord> Well, there is already code in trunk to handle what to do when a date falls on a weekend.. go to forward or back
10:40:38 <warlord> i dont think the SX code needs a complete rewrite.
10:40:45 <warlord> It's fairly robust.
10:42:04 <joslwah> It just seems extremely convoluted.
10:42:21 <joslwah> Why not be able to jump to the exact days immediately rather than jumping forward and backwards.
10:42:50 <joslwah> I may be missing something in the code. It just looks a lot more complicated than it needs to be.
10:42:52 <jsled> I don't think it needs a rewrite, considering it was basically already rewritten.
10:43:09 <jsled> The time-based stuff in particular was wholly rewritten.
10:43:21 <jsled> "jumping forward and backward" ?
10:43:35 <joslwah> O.k. I just couldn't see how to easily update it so that you could say "three business days after date x"
10:43:55 <joslwah> Yeah. To find the "next occurance" it jumps a period forward and then wants to rewind until it finds the right date.
10:44:03 <joslwah> At least that way my reading from the comments.
10:44:12 <jsled> warlord: more generally don't care, so it's not "too busy", more "can't be arsed". And if I did spend time on gnucash, that issue would be pretty far down prioirity-wise.
10:45:24 <jsled> joslwah: that doesn't sound familiar.
10:45:36 <jsled> I recall that it walks forward, but not back.
10:46:56 <warlord> maybe I should be asking chris? ;)
10:47:15 <warlord> basically, the M&L Druid just doesn't work in trunk. The SXes it creates "never fire".
10:48:49 <joslwah> jsled, Lines 155,6 of Recurrence.c
10:49:30 <joslwah> Or has that been superceeded?
10:49:42 <warlord> Recurrence is the new code. FreqSpec is (was?) the old code.
10:49:45 <jsled> was.
10:50:02 <jsled> FreqSpec was replaced with Recurrence in … 2.2, iirc.
10:50:19 <jsled> if you find Recurrence overly complex (and I don't think it is), you should have seen FreqSpec! :)
10:50:45 <warlord> LOL.
10:50:51 <joslwah> Oh, I can believe it could have been worse.
10:51:10 <warlord> The Recurrence code *is* much cleaner, and I believe it can do more, too.
10:51:32 <warlord> BIAB
10:51:42 <jsled> I'd advise against re-writing Recurrence, if only because it's relatively well tested at this point. I'm pretty sure that ± N business days can either be crafted into Recurrence or added at a higher layer more easily.
10:51:57 <joslwah> My problem with it is the case where you want to offset by n working days, where n is larger than the gap between the occurances.
10:51:58 *** warlord is now known as warlord-afk
10:52:13 <joslwah> Or should I just declare that case "off-limits" and then it is fairly easy to modify.
10:52:27 <jsled> But, as I said earlier, "generally don't care", so "coder's prerogative".
10:52:50 <joslwah> It might be reasonable to say that n < 5. And to ignore extra holidays like Easter.
10:53:32 <jsled> Yeah, that might be a tricky case. Cause you either want to push future occurrences past the adjusted occur date (likely) or not, and "not" means you need to track *both* instance dates, which would be a lot of plumbing through the code.
10:54:12 <jsled> Hmm.
10:54:30 <joslwah> That was what started me thinking.
10:55:20 <jsled> ("both" being 1/ the Recurrence-computed next-instance date and 2/ the adjusted-for-business-days *actual* occurrence date, of course)
10:55:26 <joslwah> So, it can be done properly using a rewrite, or hacked by messing around with the current code.
10:55:40 <jsled> bah.
10:55:54 <jsled> I'd not frame it that way.
10:56:20 <joslwah> Exactly. Plus, you may need to offset backwards. Lets say that the recurrence occurs tomorrow, but was 4 business days afterwards, and this thing happens every two days, so by starting with today you miss some occurence.
10:56:46 <jsled> That is, I think it can be done "properly" in the existing code, too, but I've not looked into it. Maybe it would end up being a "hack".
10:56:47 <joslwah> It all becomes a horrible mess, and you can bet that any restriction we put on it is going to be objected to by someone.
10:57:29 <joslwah> One big issue is to work out what we do about non-weekend non-business days. Are they just ignored?
10:59:32 <joslwah> I have various monies that come into an account on the 3rd business day after a fixed date, every month. It is irritating to have to fix each one up every time.
11:02:02 <joslwah> I think the way to do it is to have a wrapper. One routine works out the "base" date and another then works out the effective date.
11:02:14 <joslwah> The wrapper deals with fixing everything together.
11:02:15 * jsled nods.
11:02:20 <jsled> Just do it in the layer above entireyl.
11:02:29 <jsled> Then it can omit/combine instances as need be.
11:02:55 <jsled> Plus, understanding the concepts of "business" (and "holidays") is best left out of the Recurrence code, too.
11:02:55 <joslwah> Yeah. I'm just not quite sure why the base routine is "next instance" rather than "give me all the instances between date x and date y"
11:03:16 <jsled> Well, everything effectively does the latter.
11:03:21 <jsled> Just through looping.
11:03:31 <jsled> (Or "Through just looping" :)
11:03:47 <joslwah> So, you'd suggest having a new set of routines that use Recurrence, and then have everything migrate eventually?
11:04:23 <joslwah> Yeah, but it could be a lot more efficient. If we stored extra information in the Recurrence data structure we should be able to just generate the dates immediately.
11:04:32 <jsled> effectively yes, I suppose.
11:04:46 <jsled> I mean, I hadn't thought about it that way, but that is what would happen.
11:05:03 <jsled> The SX model would want ot use that layer rather than the recurrences directly.
11:05:14 <jsled> Or maybe they're more peers, rather than layers.
11:05:38 <jsled> the SX code creates an instance list through the Recurrences, then hands that instance list to the "business/holiday adjuster" to modify it.
11:05:54 <joslwah> So we'd need a list of country specific holidays as well.
11:06:10 <jsled> less coupling if it's a layering thing, at the cost of a wider API.
11:06:34 <jsled> maybe. Maybe don't need to start with the holiday thing, but it's worth thinking about.
11:06:42 <jsled> Coder's prerogative, again. :)
11:06:54 <joslwah> The problem with that idea is that what is relevant datewise may be different. So the SX code may select out some dates, and then have them modified, but then miss a recurrence because it wasn't in the date-range.
11:07:02 <joslwah> I think it wants to be a layer thing.
11:07:38 <joslwah> So that the modifying code can extend the date-ranges with the knowledge of the wanted offsets.
11:08:25 <jsled> There's related gnucash-devel traffic right now on this topic, btw. I don't know if you're subscribed...
11:09:17 <joslwah> I am, and was watching it.
11:09:35 <jsled> cool.
11:09:45 <joslwah> I think that an approximate solution, that doesn't miss anything, but may give the wrong date, is better than nothing.
11:09:59 <joslwah> But if it misses items then it is much worse than nothing.
11:10:32 <joslwah> But there would need to be a big warning over it that "you can't trust it because it may miss bank holidays in your country of residence" or something.
11:16:27 *** Marquel has joined #gnucash
11:16:32 <Marquel> morning.
11:18:43 <jsled> hello.
11:19:22 <Marquel> tiny little problem (as usual): i have a bill from one of my vendors. unfortunately that bill was wrong but is already booked in gnucash. now i want to enter the cancellation, but gnucash tells me i "can not book an invoice with negative balance"... :( now what should i do?
11:31:44 *** IanL has quit IRC
11:37:10 <Marquel> no idea anyone?
11:45:55 *** Rolf1 has quit IRC
12:01:08 *** warlord-afk is now known as warlord
12:02:33 <Marquel> morning warlord
12:06:27 <warlord> unpost the old bill, change it, and repost it? Or book the adjustment as a "Payment"
12:07:07 <warlord> Well, test-lots works in 2.2.8 revision. So now I get the search between 2.2.8 and current 2.2 to see where it broke.
12:13:24 <Marquel> warlord: my vendor sent a bill with negative balance, i'd rather not unpost/repost the old bill. payment is not that simple since it's a full cancellation (including all taxes). so maybe i should make this a feature request.
12:16:57 <warlord> It's been requested. I've closed every request as WONTFIX
12:17:11 <warlord> There is an open request for a "Credit Note"
12:17:42 <warlord> You have three ways to do it now:
12:17:50 <warlord> 1) Unpost, correct, and repost the old bill
12:17:53 <warlord> 2) Use the Process Payment
12:18:02 <warlord> 3) Wait for the next bill to come in and combine this one with that one.
12:22:08 <Marquel> i take step 2. 3 is a too long term operation and 1 may be ... disliked by fiscal authorities...
12:23:49 <warlord> ok
12:29:50 <Marquel> those fiscal authorities are the reason why i asked for that "credit note" ;)
12:47:36 <warlord> Okay, test-lots broke in 2.2 somewhere between r17857 and r17864... Getting closer to tracking it down.
12:48:24 *** andyt has joined #gnucash
12:56:34 <warlord> Still broken in r17861
12:57:01 <warlord> Er.. oops...
12:57:06 *** ErKa has joined #gnucash
12:57:21 <warlord> pulled out the wrong revision in that test.
13:07:45 <warlord> Okay, the break is after r17861
13:11:15 *** palatin has joined #gnucash
13:25:01 <warlord> still broken in r17863
13:25:29 <warlord> now testing r17862. So the break is either r17862 or r17863
13:26:49 *** ErKa has quit IRC
13:26:53 *** Zoolooc_ has joined #gnucash
13:28:50 *** MechtiIde has quit IRC
13:45:37 <ska> bugzilla.gnome.org is extrementy slow.
13:45:55 <warlord> maybe it just doesn't like you?
13:45:58 <warlord> (we don't control that server)
13:46:55 <jsled> seems reasonably responsive here.
13:47:43 <warlord> Okay, r17862 broke test-lots.
13:48:56 <warlord> I dont understand /how/ it broke test-lots... but it did.
13:55:50 <warlord> now.. why is it working in trunk but not 2.2? *ponders*
13:58:14 *** hvx has joined #gnucash
14:03:02 <Marquel> warlord: i know that windows (well, nt4) can be posessed by ... say devil - and thus made working by using crosses... but bugzilla? that's somehow beyond my belief ;)
14:06:18 <warlord> heh
14:07:18 <Marquel> i know it's hard to believe, but i've experienced it myself...
14:07:43 *** bentob0x has quit IRC
14:11:30 <goodger> Marquel: I suggest you threaten it
14:11:41 <goodger> attaching a Trac tarball to a bug report
14:15:23 <Marquel> goodger: i could also give it some patches i wrote for trac... ;)
14:16:56 <goodger> mmm, would show you mean business
14:18:04 <Marquel> goodger: difference might be: a copy of opensource code is not as worthy to me as was my cross (which is why it was decided to scrub the box and go for winnt5.... then...) *g*
14:19:05 <goodger> I see
14:20:15 <Marquel> *sigh* the good ol'days... when "plug'n'pray" was working... now it is "plug'n'fixit"...
14:20:51 <Marquel> anyway it was month's end and i need some paper trails. bbl.
14:20:56 <goodger> yeah, prayer has been deprecated ^_^
14:20:58 <goodger> bye
14:23:53 <warlord> heheh
14:25:22 <goodger> <texan>where I come from, we get of our asses and _do_ something!</texan>
14:25:33 <warlord> LOL.
14:25:41 <goodger> *off
14:25:55 <warlord> Hey, I just tracked down the test-lots failure in 2.2 to r17862
14:25:58 <warlord> THat's doing something.
14:26:10 <warlord> However I dont understand WHY that breaks the test in 2.2 but not trunk.
14:30:40 *** MechtiIde has joined #gnucash
14:37:09 *** eukreign has quit IRC
15:25:22 *** sjc has joined #gnucash
15:26:04 *** hvx has quit IRC
15:55:02 *** |gunni| has joined #gnucash
15:56:28 *** |gunni| has quit IRC
16:26:08 *** cortana has joined #gnucash
16:30:14 <Marquel> gn8
16:30:26 *** Marquel has quit IRC
16:45:21 *** warlord has quit IRC
16:56:28 *** warlord has joined #gnucash
16:56:29 *** gncbot sets mode: +o warlord
16:58:22 *** warlord has quit IRC
16:58:44 *** warlord has joined #gnucash
16:58:44 *** gncbot sets mode: +o warlord
17:36:50 *** palatin has quit IRC
17:39:43 *** nomeata has quit IRC
18:05:35 *** warlord has quit IRC
18:17:26 *** andyt has quit IRC
18:23:09 *** warlord has joined #gnucash
18:23:09 *** gncbot sets mode: +o warlord
18:58:14 <goodger> hu
18:58:15 <goodger> *hi
18:58:24 <goodger> is it possible to reorder transactions that occurred on the same day?
19:01:20 <warlord> Default order is Post Date, Check#, Date Entered.
19:01:36 <warlord> So.. You could re-enter them to change the date-entered value to reorder them.
19:01:40 <warlord> Or you could use the check#
19:02:06 *** Zoolooc_ has quit IRC
19:07:22 *** aindilis has joined #gnucash
19:14:00 <goodger> I tried that
19:14:10 <goodger> re-entering them, rather
19:14:19 <goodger> in order
19:14:24 <goodger> they reordered themselves
19:14:42 <goodger> and there was no associated cheque number unfortunately
19:27:34 <warlord> are some debits and some credits?
19:27:52 <warlord> And even if there isn't an associated check#, you can just put something into the Num column anyways.
20:12:57 *** warlord is now known as warlord-afk
20:22:14 *** twunder has joined #gnucash
20:24:14 *** dissident has quit IRC
20:36:25 *** Rolf has joined #gnucash
20:40:00 *** aindilis has quit IRC
20:49:23 *** cortana has quit IRC
20:52:35 *** sjc has quit IRC
21:34:42 *** amitz has joined #gnucash
22:15:08 *** twunder has quit IRC
22:20:30 <amitz> where can I get comprehensive list of the "small business feature" of gnucash 2.2.8?
22:36:56 *** twunder has joined #gnucash
22:37:12 *** Rolf has quit IRC
22:37:16 *** twunder has quit IRC
22:38:49 *** warlord-afk is now known as warlord
22:38:57 *** Rolf has joined #gnucash
22:39:01 <warlord> amitz: I'm not sure there's a list per se... It's A/R, A/P, etc.
22:43:51 *** Rolf has quit IRC
22:48:41 *** draemon has quit IRC
22:53:23 <amitz> warlord, hmm, if I manage to compile one, where can I submit such list?
22:53:42 <warlord> the wiki?
22:54:07 <amitz> hmm need to register... ok.
22:58:57 *** Rolf has joined #gnucash
23:05:47 *** Rolf has quit IRC
23:06:30 *** Rolf has joined #gnucash
23:51:10 *** bobnormal has joined #gnucash
23:52:12 <bobnormal> hey there, i'm implementing an automated accounting solution for a service business and i'm making a design decision regarding data storage (sql based). is it really important to keep receivable/payable for a given account totally separate? or would splitting by positive/negative value and storing together be fine
23:52:15 <bobnormal> im leaning towards the latter
23:52:37 <bobnormal> but i'm no accounting expert .. maybe someone with more experience has some insight
23:56:10 <warlord> What do you mean by "a given account"
23:56:20 <bobnormal> well i'm managing multiple accounts
23:56:32 <warlord> Define "account".
23:56:40 <warlord> Do you mean "Customer" or "Vendor"?
23:56:45 <warlord> (to use Gnucash terminology)
23:57:13 <bobnormal> in this software it's 'a collection of data regarding past and future transactions that is associated with a real bank account and business, and in a single currency'
23:57:29 <bobnormal> neither i'm just in here because i know gnucash is one of the most popular open source accounting packages
23:58:33 <warlord> Okay, so you are actually talking about a Bank Account... In which case I dont understand what you mean by receivable/payable in this context.
23:58:42 <bobnormal> 'expected future transactions'
23:58:46 <warlord> Yes, you want to keep the debits and credits against that account in one place.
23:58:57 <warlord> .. and yes, you want to store them together.
23:59:25 <bobnormal> ok. figured as much. i dont understand why there's such an emphasis on 'accounts receivable' and 'accounts payable' as separate entities in regular accounting terminology
23:59:40 <bobnormal> i guess its just a hangover from paper accounting and it should just be 'future transactions'