2008-09-16 GnuCash IRC logs
00:22:42 *** rgotten has joined #gnucash
00:24:06 <rgotten> i am trying to create a new account and it comes to the placeholder i wnat to checkmark some options but it does not let me do this, any idea why?
00:36:17 *** rgotten has quit IRC
00:37:49 <jsled> gncbot: tell rgotten, you should stick around for more than 13 minutes. Otherwise, use the mailing list.
00:37:49 <gncbot> jsled: The operation succeeded.
00:41:15 *** rgotten has joined #gnucash
00:42:07 <rgotten> Any body can help me, why can not chek mark a placeholder?
00:54:04 *** rgotten has left #gnucash
01:56:44 *** Toba has joined #gnucash
02:01:03 *** esperegu has joined #gnucash
03:21:32 *** ErKa has joined #gnucash
06:11:46 *** ErKa has quit IRC
07:31:02 *** esperegu has quit IRC
07:31:04 *** nawab has quit IRC
07:31:12 *** esperegu has joined #gnucash
07:31:13 *** nawab has joined #gnucash
07:39:16 *** andi5 has joined #gnucash
07:39:16 *** gncbot sets mode: +o andi5
07:50:44 *** twunder has joined #gnucash
07:58:25 *** warlord-afk is now known as warlord
07:59:07 <warlord> @tell rgotten you should stick around for more than 13 minutes. Otherwise, use the mailing list.
07:59:07 <gncbot> warlord: The operation succeeded.
08:07:47 *** JimRaehl has left #gnucash
08:16:48 *** ErKa has joined #gnucash
08:43:55 *** twunder has quit IRC
08:46:44 *** kielein has joined #gnucash
09:02:39 *** ErKa has quit IRC
09:19:54 *** ErKa has joined #gnucash
09:37:01 *** ben_goodger has joined #gnucash
10:21:49 *** ben_goodger has quit IRC
10:32:33 *** ben_goodger has joined #gnucash
10:51:25 *** ErKa has quit IRC
11:43:56 *** andrewsw has quit IRC
11:47:11 *** andrewsw has joined #gnucash
11:47:11 *** gncbot sets mode: +o andrewsw
12:18:03 *** ErKa has joined #gnucash
12:26:35 <warlord> gnucash ;)
12:26:57 <warlord> test-lots
12:27:21 <andi5> right
12:27:31 <andi5> it has become slower on trunk by magnitudes
12:28:26 <andi5> i just hope it has nothing to do with pending backports ;-) [moving slowly forward]
12:30:22 <warlord> It could be Phil's changes?
12:31:02 <andi5> yep
12:31:10 <andi5> but i just do not know
12:31:21 <warlord> I'm not sure how to read the graph
12:32:02 <andi5> it has been produced by gprof2dot and oprofile, just as the example on http://code.google.com/p/jrfonseca/wiki/Gprof2Dot is written
12:32:31 <andi5> and yes, it takes some time to read it
12:35:39 <andi5> i cannot describe right now... have to leave for a while
12:37:18 <warlord> Okay.
12:37:25 <warlord> It's lunchtime for me.
12:37:48 <andi5> great, see you
12:38:00 *** andi5 has quit IRC
12:38:05 <Toba> dang andrewsw
12:38:15 <Toba> er, andi
12:38:19 <Toba> that's some crazy oprofile
12:40:32 <warlord> 23% time spent in gnc_numeric_mul!
12:40:56 <Toba> can't overflow
12:40:58 <Toba> it makes sense
12:58:57 <warlord> It's still a long time in that function.
13:08:29 *** donfede has joined #gnucash
13:26:10 *** warlord is now known as warlord-afk
13:56:51 *** nelt has quit IRC
14:03:10 *** sjc has joined #gnucash
14:32:14 *** jsled has quit IRC
14:32:15 *** ErKa has quit IRC
14:37:00 *** ben_goodger has quit IRC
14:38:17 *** twunder has joined #gnucash
14:39:38 *** ben_goodger has joined #gnucash
14:54:56 *** warlord-afk is now known as warlord
14:57:49 *** bandersnatch has joined #gnucash
14:58:09 <bandersnatch> Hello and a good day
14:58:39 <bandersnatch> I have an scm script for gnucash that ran under 2.0
14:59:45 <bandersnatch> with the newer version 2.2 this breaks on a couple of items. i have replaced some calls from gnc:* to gnc-*
15:00:04 <bandersnatch> I have got as far as having the script load and create its menu item
15:01:08 <bandersnatch> running the functionality currently breaks on gnc:get-current-session ( and probably some other from the session family )
15:01:58 <bandersnatch> I have not found a replacement to this in any of the other *scm files. ( in price-quotes.scm was an example but this vanished in the current version.
15:02:32 <bandersnatch> Any good tips on how to proceed from here? ( I am not really at home with guile )
15:07:00 <warlord> bandersnatch: most of the GnuCash APIs were converted over to their C counterpart. So find the C API that matches the API you wanted.
15:07:13 <warlord> (see -- this is why you should have donated your report)
15:08:30 *** ben_goodger has quit IRC
15:09:58 <bandersnatch> Thanks, which report?
15:12:09 *** ben_goodger has joined #gnucash
15:16:16 <bandersnatch> did you mean the abort backtrace?
15:16:23 *** |gunni| has joined #gnucash
15:17:21 <warlord> No, I mean the report that you've written.
15:17:32 *** jsled has joined #gnucash
15:17:32 *** gncbot sets mode: +o jsled
15:17:43 <warlord> (I was assuming that your scheme file was a report)
15:19:32 <bandersnatch> The file is ibr-gnc-module.scm which provides a variety of functions . I use the export to csv functionality.
15:20:51 <bandersnatch> This is not my script but I have made a couple of small adjustments to the exported items. The rest is a big dark gray box at the moment.
15:22:10 <warlord> Huh. Never heard of that.
15:25:47 <bandersnatch> this is athe url of a later version : http://www.ibrieger.de/misc/linux/gnucash/ibr-gnc-module.scm.gz
15:26:20 <bandersnatch> just found it ( going to compare against my version )
15:27:54 <warlord> ok
15:29:44 <bandersnatch> well, it is still based in the old version with a lot of small changes in functionality. ( but no fixes towards 2.2 compat )
15:32:33 <warlord> Well, I recommend you contact the original author.
15:34:37 *** hjb has joined #gnucash
15:35:03 <hjb> Hi. Simple question but not answered by the tutorial (as far as i have worked through it):
15:35:12 <hjb> What does the "close book" tool do?
15:35:59 <bandersnatch> Right, i already did that. no answer yet. I had thought it _easier_ to fix ( and was rather happy to have at least load without fireworks ).
15:36:37 <warlord> hjb: zeroizes your Income and Expense accounts.
15:37:06 <warlord> bandersnatch: fair enough.
15:37:20 <hjb> warlord and by that let my equity reflect the current state?
15:38:09 <warlord> hjb: effectively, yes.
15:39:14 <hjb> warlord Ok, thanks :-) Than i've got that right
15:39:42 <warlord> ok
15:40:22 <hjb> Should be documented ;-)
15:40:28 <warlord> send in a patch
15:40:33 <hjb> Ok :-)
15:40:57 <warlord> (is it not on the wiki/FAQ?)
15:41:10 <hjb> (maybe ...) :-/
15:41:18 <hjb> sorry for bothering you with it, than ...
15:42:56 *** hjb has quit IRC
15:51:02 *** bandersnatch has quit IRC
16:04:27 *** bandersnatch has joined #gnucash
16:04:45 <bandersnatch> got dropped out.
16:05:11 <bandersnatch> thanks for your help warlord.
16:06:14 <warlord> you're weclome
16:15:14 *** bandersnatch has quit IRC
16:29:40 *** ErKa has joined #gnucash
16:34:09 *** twunder has quit IRC
16:34:46 *** twunder has joined #gnucash
16:43:34 *** HM2K has joined #gnucash
16:43:42 *** ErKa has quit IRC
16:43:49 <HM2K> has anyone ever successfully migrated from sage to gnucash?
16:47:48 *** esperegu has quit IRC
16:54:45 *** ErKa has joined #gnucash
17:07:17 *** twunder has quit IRC
17:15:29 <warlord> HM2K: you might try asking on the -user list. Can you export your SAGE txns as QIF?
17:16:24 <HM2K> not sure, SAGE is the leading accountancy software here in the UK, it's fairly tight, but there is an API
17:37:25 *** cedayiv has joined #gnucash
17:56:13 *** ErKa has quit IRC
18:00:31 *** |gunni| has quit IRC
18:34:13 *** kielein has quit IRC
18:35:38 *** andi5 has joined #gnucash
18:35:39 *** gncbot sets mode: +o andi5
18:39:02 *** warlord is now known as warlord-afk
18:39:21 <andi5> someone should aquaint gncbot with cedayiv :)
18:39:40 <cedayiv> say what?
18:40:05 <andi5> you get a yellow start, just as the one i finaly got mailed from ebay for 10 auctions :)
18:40:10 <andi5> s,start,star,
18:40:25 <jsled> @op cedayiv
18:40:26 *** gncbot sets mode: +o cedayiv
18:40:37 <andi5> does that last for a while?
18:41:10 <jsled> you should "/msg gncbot list" and "/msg gncbot help" to figure out how to identify with gncbot. Then we can give the chanop token so you'll get op'ed on channel entry.
18:41:25 <jsled> andi5: only until /part.
18:41:56 <cedayiv> I have a yellow ebay star too
18:42:00 <andi5> can everyone identify with gncbot and get op'ed?
18:42:02 <andi5> hm :)
18:42:06 <andi5> yeehaw :)
18:42:54 <andi5> cedayiv: btw, the the patch you sent was exactly how i would have solved it.... the question still remains whether that is actually a problem
18:43:08 <andi5> have you ever seen those numbers anywhere?
18:43:36 <cedayiv> no, that's why I wondered how to test them
18:44:11 <andi5> oh, on linux you can gdb-attach to a running gnucash process and simply run a few tests
18:44:27 <andi5> or you can integrate tests with our test suite (which should work on win32 as well)
18:45:42 <andi5> cedayiv: that is actually an interesting question whose answer i forgot: can you run gdb and issue "attach $pid", where pid is some process identifier of the gnucash process?
18:46:28 <andi5> i guess the problem is resolving symbols, so a few "dir" commands will probably be necessary, maybe also other stuff, like "at address x there is library l from fn" ...
18:46:57 <cedayiv> I can't attach on winxp/mingw. It denies me.
18:47:33 <andi5> you do not need msys or mingw for gdb... it even runs better when started from cmd.exe or the explorer
18:47:36 <cedayiv> What am I supposed to say to gncbot?
18:47:48 <andi5> then you can actually interrupt processes without killing them ;-)
18:47:55 <andi5> help
18:48:21 <cedayiv> it just says "(help [<plugin>] [<command>])". not very helpful
18:48:27 <andi5> help list then :)
18:48:47 <andi5> i mean list
18:49:46 <cedayiv> It says "Admin, AutoMode, Channel, ChannelLogger, Config, Later, Misc, Owner, Seen, and User"
18:51:04 <andi5> list user ... use register and afterwards identify
18:51:29 <cedayiv> I would like to be able to attach to a running gnucash, but since I can't, do you know how to stop it while running? Set a breakpoint somewhere (where?) beforehand?
18:52:10 <andi5> i know how to start it :)
18:53:04 <andi5> add gdb in gnucash.bat before gnucash, then run it.... at the gdb prompt add a breakpoint "b load_gnucash_modules", run "r", finish that function "fin", then add breakpoints "b...", continue "c"
18:53:12 <andi5> that is what i used to do
18:53:57 <andi5> how to stop it.... i could try a little bit
18:54:03 <cedayiv> OK, the "load_gnucash_modules" is the important bit that I didn't know, thanks.
18:54:33 <andi5> well, before that function the list of possible breakpoints is rather limited, afterwards you can break at almost every position
18:55:17 <cedayiv> So for gncbot I should use "register cedayiv" and specify a password?
18:55:26 <andi5> i think so
18:55:43 <andi5> i could check my old gncbot logs :)
18:56:55 <andi5> i am sorry, even the first log entry from 2006 starts with "identify..." ... succeeded ;-)
18:57:15 <cedayiv> I did the "register" and "identify". It seemed to "succeed".
18:57:29 <andi5> you could try to part and come back
18:58:41 <cedayiv> Do I need to do something with the hostmask? I am on DSL so my hostname could change every day.
18:59:17 <andi5> i do not know... it just works for me
18:59:22 <andi5> and i have dsl as well
19:00:21 <cedayiv> OK, I'll try it.
19:00:39 *** cedayiv has left #gnucash
19:00:40 <andi5> cedayiv: are r17533 and r17534 ok for you? ... btw, it would be nice if someone reviewed my (few) commits :)
19:00:43 <andi5> ahhh
19:01:17 *** cedayiv has joined #gnucash
19:01:28 <andi5> [OT] should i install the sp3 in my virtual windows?
19:02:21 <cedayiv> I installed SP3. I didn't notice any difference before/after.
19:02:38 <andi5> hm... ok... then i will defer that update
19:02:40 <cedayiv> My WinXP isn't virtual though.
19:02:47 <jsled> cedayiv: try @op'in.
19:02:52 <jsled> @op 'ing, even.
19:02:52 <gncbot> jsled: Error: 'ing, is not in #gnucash.
19:02:59 <jsled> gncbot: :P
19:02:59 <gncbot> jsled: Error: ":P" is not a valid command.
19:03:06 <andi5> @die
19:03:06 <gncbot> andi5: Error: "die" is not a valid command.
19:03:16 <cedayiv> @op
19:03:16 *** gncbot sets mode: +o cedayiv
19:03:26 <andi5> what he actually wanted to say "your wish is my command"
19:03:32 <jsled> should happen on channel join / ident, now.
19:03:57 <andi5> [2nd try] cedayiv: are r17533 and r17534 ok for you? ... btw, it would be nice if someone reviewed my (few) commits :)
19:05:50 <cedayiv> Have not tried them yet. What is G_LIKELY?
19:06:10 <jsled> cedayiv: It's a branch-prediction hint.
19:06:36 <jsled> Or, so it claims. I'm curious how it's implemented, actaully.
19:06:48 * jsled used it somewhat blindly in the qof logging stuff.
19:06:58 <andi5> basically it is only a hint for the optimizer part of the compiler
19:07:20 <cedayiv> Huh. I have always just tried to arrange branches in order of likelyhood. That's nice.
19:07:45 <andi5> well, it should only be used if the first branch is really almost always the right one...
19:07:48 <jsled> [[[
19:07:48 <jsled> #define G_LIKELY(expr) (__builtin_expect (_G_BOOLEAN_EXPR(expr), 1))
19:07:48 <jsled> #define G_UNLIKELY(expr) (__builtin_expect (_G_BOOLEAN_EXPR(expr), 0))
19:07:49 <jsled> ]]]
19:07:53 <andi5> the alternative becomes expensive, iirc
19:08:25 <jsled> (/usr/include/glib-2.0/glib/gmacros.h) Conditional on gcc > 2 and "__optimize__"
19:09:16 <cedayiv> jsled: Do I need to check the "Enable Identification Server during connection process" option in my Chatzilla?
19:09:54 <jsled> I don't believe so ... I've always thought that was some sort of ircd-level credentialing.
19:10:12 <jsled> Oh, or maybe it's the identd callback stuff.
19:10:15 <cedayiv> ??
19:10:30 <andi5> is that a public "open" identd?
19:10:37 <cedayiv> I'll try a complete disconnect and let's see what happens
19:10:54 *** cedayiv has quit IRC
19:11:06 <andi5> or does it start an own identd on the host? :)
19:11:35 *** cedayiv has joined #gnucash
19:11:36 *** gncbot sets mode: +o cedayiv
19:11:47 <andi5> auto star mode
19:11:53 <cedayiv> Cool, it worked.
19:11:56 <jsled> maybe. /me notes the second ghit for "identd" is "identd (identification protocol) is pointless and potentially dangerous" -- <http://www.clock.org/~fair/opinion/identd.html>, which always seemed to be my impression.
19:12:17 <andi5> indeed
19:14:03 <cedayiv> Now I'll have to find out what an "op" does... (complete IRC noob)
19:14:12 <andi5> lol
19:14:18 <andi5> yeah, let us find a lurker and kick it
19:15:52 <cedayiv> I'm glad to see r17533. That was on my list of things to do. There is a bug # for it.
19:16:16 <andi5> oh, you know those bugs better than i could search them :)
19:16:26 <andi5> not that i actually did the searching...
19:16:55 <cedayiv> (looking)
19:17:44 <cedayiv> Bug 430121
19:18:40 <andi5> thanks
19:18:53 <cedayiv> Not sure if you also fixed bug 518733 in r17527. (Haven't tested for that.)
19:20:17 <andi5> no, i think that is another bug that has been fixed in gnucash 2.2.6 or 2.2.5
19:20:47 <andi5> [2.2.6] Correct quickfill directions in transfer dialogs
19:21:44 <cedayiv> OK, guess it can be closed then. I only read it - I never even got around to trying to reproduce it.
19:22:59 <andi5> done
19:23:40 <cedayiv> Great
19:24:15 <andi5> ok, gdb can attach to running gnucash installation (interrupting the process as intended)
19:24:23 <andi5> but the stacktrace is not developer-friendly
19:26:15 <andi5> cedayiv: maybe you find a better solution than that in the two mails linked to in http://wiki.gnucash.org/wiki/Windows#Debugging_with_gdb
19:26:16 <cedayiv> I have a patch for bug 339290 but not tested yet... not enough time in the day!
19:27:40 <andi5> cedayiv: the more i looked at the register code, i observed that multicurrency support must be really bad
19:28:00 <andi5> the prices are printed in all sorts of weird print infos and fractions
19:28:18 <cedayiv> Yes. Not nice.
19:28:41 <cedayiv> For the exchange dialog I prefer to see the fractions though
19:28:55 <andi5> either someone really did not understand what he was what the register is supposed to show, or exactly that changed over time
19:30:10 <cedayiv> There is another problem with print infos that I don't remember right now... I will have to think about it.
19:30:44 <cedayiv> I have been starting gnucash with exec gdb --args gnucash-bin "$@"
19:30:47 <andi5> in the hierarchy druid
19:31:17 <cedayiv> ... will have to try exec start gdb gnucash-bin "$@"
19:31:51 <cedayiv> What does "start" do?
19:32:25 <andi5> what about gdb --args "$@" gnucash-bin?
19:33:09 <andi5> i am not sure... it works ;-)
19:33:26 <cedayiv> good enough
19:33:32 <andi5> i think otherwise command windows would not be closed, or something like that?
19:34:31 <andi5> btw, i think you mean %*, right?
19:35:00 <cedayiv> Oh... well I always start from a mingw shell so I don't care about command windows
19:35:12 <andi5> ah, right
19:35:54 <andi5> maybe we can overhaul all these "files are retrieved from windows installation instead" and "paths are hard-coded in registry" stuff
19:36:00 <andi5> once aqbanking3 compiles on windows
19:36:08 <andi5> i mean, we can use that by default
19:36:25 <cedayiv> The print info thing (I remember now) is that there is currently no way to print non-decimal numbers except as fractions. We would need to add a "force_decimal" option to the print info.
19:36:32 <cedayiv> See: https://lists.gnucash.org/pipermail/gnucash-devel/2008-August/023697.html
19:36:46 *** kimmo has quit IRC
19:36:49 <andi5> i still think there is an issue in the hierarchy druid :)
19:37:27 <andi5> you would use that in exchange rate fields?
19:37:30 <cedayiv> Which one is that? The one that makes an initial account tree?
19:37:36 <andi5> yep
19:38:35 <cedayiv> I don't know that we actually need a "force_decimal" option at all. I'm just saying that currently 2/3 can only be printed as "2/3" and never as "0.667"
19:40:10 <cedayiv> So far that doesn't seem to have caused any problems. But then, prices were always getting rounded before.
19:40:25 <andi5> force_fit -1?
19:40:48 <andi5> but it is counter-intuiute
19:40:51 <andi5> intuitive...
19:41:17 <andi5> \infty would be better :)
19:42:31 <cedayiv> If you don't want to add a "force_decimal", then you would need four options for "force_fit". See the last table in that email.
19:42:58 <andi5> yes
19:43:04 <cedayiv> Seems easier to just add a "force_decimal" boolean.
19:43:41 <andi5> i was ignorant, those flags are neither booleans nor uints, simply a bit flags
19:45:27 <cedayiv> I guess we will see if anyone complains about printing "2/3". I think prices might be the only thing that can be fractional now anyway.
19:46:25 <cedayiv> I will test your latest patches when I get a chance. Probably later today.
19:46:35 <cedayiv> I have to go now.
19:46:41 <andi5> ok, cya
19:46:54 <cedayiv> l8r
19:47:05 *** cedayiv has quit IRC
20:39:41 *** JimRaehl has joined #gnucash
21:05:38 *** sjc has quit IRC
21:42:16 *** BlackBsd has joined #GnuCash
21:42:30 *** BlackBsd has quit IRC
21:48:04 *** jcreigh has joined #gnucash