2008-09-18 GnuCash IRC logs
02:12:08 *** Pupeno has joined #gnucash
02:12:17 <Pupeno> Hello.
02:12:58 <Pupeno> I've just put a transaction from an EUR account to a CHF account, and it's just removing the same number from EUR and adding it as CHF, without letting me specify the exchange rate.
02:13:34 <Pupeno> How do I specify the exchange rate?
02:43:38 *** esperegu has joined #gnucash
03:16:59 *** cedayiv-afk is now known as cedayiv
03:20:53 <cedayiv> There are a number of bugs related to transferring between accounts with different currencies. The 2.2.7 release (due any day now) fixes a bunch of them.
03:27:55 <cedayiv> The exchange rate dialog should pop up when you enter a new transaction. However, if you go back to an old transaction and change one of the splits then the exchange rate dialog does not show. You have to click Actions->Edit Exchange Rate.
03:28:15 <cedayiv> Hope this helps.
03:29:40 *** ErKa has joined #gnucash
04:20:55 * Rolf1 curses gnucash's build system
04:21:52 *** martincleaver has joined #gnucash
04:48:02 *** cedayiv has quit IRC
04:57:53 *** ecraven has joined #gnucash
04:57:57 <ecraven> hello ;)
04:58:12 <ecraven> does gnucash support batch editing? i want to set the target of 10 transactions to Income:Bank
05:13:47 *** Pupeno has quit IRC
05:22:46 *** Pupeno has joined #gnucash
05:46:38 *** kielein has joined #gnucash
05:54:12 *** ErKa has quit IRC
05:58:38 <Rolf1> ecraven: Maybe you want a single split transaction?
05:59:37 *** JimRaehl has quit IRC
06:10:07 *** ErKa has joined #gnucash
06:13:55 *** Pupeno has quit IRC
06:14:09 <Rolf1> warlord-afk: dpkg-source cannot be forced to include the .git directory and I think it is right in doing that
06:14:44 <Rolf1> I am now trying to beat the assumptions the gnucash build system is making into shape and I hope I can get some help along the way
06:15:02 <Rolf1> I have forced detection to svn build at configure time via http://rafb.net/p/DITBM746.html
06:15:25 <Rolf1> But apparently there is more
06:16:02 <Rolf1> http://rafb.net/p/Cl5On186.html
06:18:42 <Rolf1> warlord-afk: what is it expecting that isn't there?
06:23:32 *** Pupeno has joined #gnucash
06:35:31 <ecraven> Rolf1: no, i have 12 transactions over an entire year, paying the rent. now i want to change all of them to go to account Expenses:Rent
06:51:36 *** Admin has joined #gnucash
06:51:57 <Admin> hello room
06:52:05 *** Admin is now known as page404
06:52:22 <page404> !command
06:53:43 <page404> I want to ask something about expense voucher. i think i have found a bug on gnucash
06:54:06 <page404> does anyone can help me?
06:57:48 <page404> ok here is the bug
07:00:03 <page404> when i post the invoice i cannot edit post date or the due date?does anyone know about this?
07:04:56 *** twunder has joined #gnucash
07:19:27 *** page404 has quit IRC
07:33:03 *** aindilis` has quit IRC
07:46:54 *** twunder has quit IRC
08:22:34 *** warlord-afk is now known as warlord
08:22:52 <warlord> jsled: FYI, you do not need to rebuild gnucash after installing F::Q
08:23:02 *** twunder has joined #gnucash
08:25:24 <warlord> Rolf1: the gnc version header file.
08:26:03 <warlord> And no, I think dpkg-source is NOT doing the right thing..
08:26:59 <warlord> ecraven: You cannot operate on multiple transactions at once. BUt you can perform a "Find Transactions" operation to list only those 12 (assuming there are some criteria you have to find them easily) and then modify them one by one.
08:27:09 <Rolf1> you are the only one in the world to think so
08:27:35 <warlord> Wouldn't be the first time that the rest of the world is wrong ;)
08:28:17 <Rolf1> While that is a possibility, the chances of the single person being overconfident are much greater
08:28:25 <ecraven> warlord: which is a lot of work... any chance of those operations on multiple transactions at once being implemented?
08:28:28 <warlord> * shrugs *
08:28:30 <Rolf1> everybody I talked to described the gnucash build system as broken
08:28:51 <warlord> Bring them here and I'll explain what it's doing.
08:29:11 <Rolf1> gnc-svninfo.h ?
08:29:18 <Rolf1> Is that what it is?
08:30:23 <warlord> gnc-version.h AND gnc-svninfo.h
08:30:49 <Rolf1> how are they being created?
08:30:58 <warlord> RTFMf
08:32:38 <warlord> (Mf == Makefile)
08:34:03 <Rolf1> nothing about that in there
08:34:31 <warlord> Then you're blind
08:34:35 <warlord> (or worse.)
08:34:50 <warlord> Which Makefile are you looking at?
08:34:56 *** esperegu has quit IRC
08:35:10 <Rolf1> Of course the top-level
08:35:13 <Rolf1> But I found it now
08:38:36 <warlord> okay
08:38:42 <warlord> glad you found it
08:44:53 *** ecraven has left #gnucash
09:00:44 *** esperegu has joined #gnucash
10:12:11 *** cedayiv-afk has joined #gnucash
10:12:42 *** cedayiv-afk is now known as cedayiv
10:14:30 <cedayiv> @op
10:14:31 *** gncbot sets mode: +o cedayiv
10:21:58 *** andi5 has joined #gnucash
10:21:58 *** gncbot sets mode: +o andi5
10:26:23 <andi5> Rolf1: it might be easier for you to just patch util/gnc-svnversion to print out a fix version number string.... and no, warlord is not the only one thinking that dpkg-source info is not doing the right thing... either you are misusing the tool for something it was not intended for, or dpkg-source removes files that are needed to build gnucash from svn/git/whatever... and: if you think gnucash's build system is broken here (i do not doubt that the
10:26:33 <andi5> s,fix,fixed,
10:27:10 <warlord> andi5: lost it at "doubt that the"
10:27:20 <andi5> (i do not doubt that there are a few possible improvements, actually), then you please come up with a real world scenario the current detection of version info is behaving badly and needs patching
10:27:54 <andi5> damn, i tend to forget to remove single words when correcting my first try to translate :)
10:28:04 <Rolf1> andi5: I actually did patch gnc-svnversion that way
10:28:18 <Rolf1> It parses the debian changelog
10:28:23 <Rolf1> I consider that a bad hack
10:30:06 <warlord> Yeah, that's pretty obnoxious.
10:30:13 <andi5> well, i am not sure how gnucash can change anything here.... you are building from a something that neither provides a version header file nor a way to determine the version itself :)
10:30:59 <warlord> You might as well run "make dist" and build from the tarball instead of using the broken tool you're currently using.
10:32:07 <andi5> btw, AM_PYTHON_DEVEL in its current state is unusable in my distribution... it makes evil assumptions about where header files have to reside and alike.... i will have to try to update the macros files someday, maybe the newer version works better
10:33:51 <Rolf1> warlord: the tool ain't broken
10:34:08 <Rolf1> it is widely considered bad practice to include .svn or .git in a tarball
10:34:21 <Rolf1> so dpkg-source does not include it
10:34:26 <andi5> well, why is it creating a tarball at all? cannot you provide it?
10:34:43 <warlord> ==andi5 -- if you want to build a tarball use "make dist"
10:34:55 <warlord> (which wont include the .git or .svn hooks)
10:34:58 <Rolf1> (although I would like to have a way to circumvent that, because it would make it easier for me. but the tarball would be "broken")
10:35:13 <warlord> a "make dist" tarball would never be broken
10:35:19 <Rolf1> warlord: Have you ever used pbuilder?
10:35:29 <Rolf1> Do you understand why you might want to use it?
10:35:37 <andi5> or he can run tar -cjpf himself
10:35:39 <warlord> I dont do debian/ubunti
10:35:51 <warlord> andi5: true
10:35:57 <warlord> I have used mock
10:36:09 <Rolf1> Well, then you should be more careful about what you say
10:36:21 <Rolf1> pbuilder is a very nice tool for a lot of things
10:36:23 <warlord> Oh, I'm picking my words QUITE carefully
10:36:28 <andi5> :-D
10:36:29 <Rolf1> one for ensuring compilation from scratch
10:36:47 <Rolf1> two for making sure build-time dependencies don't affect the host
10:36:52 <Rolf1> and that is important for me
10:36:54 <andi5> from released tarballs, right?
10:36:59 <Rolf1> no
10:37:02 <warlord> I'm sure it's a great tool for lots of things, but you're breaking invariants here left and right by making an assumption that you can copy an svn tree without massaging it.
10:37:03 <Rolf1> pdeduild
10:37:08 <Rolf1> andi5: pdebuild
10:37:35 <andi5> so either it checks out the source itself, or it uses released tarballs, everything else looks weird imho
10:37:39 <warlord> (or, in this case, massing it inappropriately)
10:38:31 * Rolf1 has long since learned that arguing with gnucash devs is usually a waste of everybody's time
10:38:33 <Rolf1> mine included
10:38:44 <andi5> Compile as gobolinux tool does check out from repositories, if you are using an cvs/svn/whatever recipe :)
10:39:14 <andi5> well, actually i really try to understand
10:39:44 <warlord> Oh, I COMPLETELY understand the need to separate the build-time deps from your local system.
10:39:50 <warlord> s/need/want
10:41:49 <warlord> But if you're building from SVN you can't just "svn co" and then build a tarball without the .svn dirs and expect it to work.
10:42:03 <andi5> Rolf1: have you read http://workaround.org/moin/SvnBuildpackage? (i have not yet ;-))
10:42:08 <Rolf1> warlord: not in gnucash's case, no
10:42:23 <warlord> Rolf1: i would never assume it in ANY case.
10:42:28 <Rolf1> andi5: It will run into the exact same problem
10:42:29 <warlord> it's not just gnucash.
10:42:47 <andi5> oh, ok
10:42:52 <Rolf1> I have packaged a number of packages
10:43:06 <Rolf1> I have talked to a lot of other people who package for debian and ubuntu
10:43:30 <Rolf1> Everybody in that domain considers gnucash's build system broken
10:43:34 <jsled> Rolf1: then how do others include a VC version number in the sources without a valid checkout?
10:43:36 <Rolf1> I know what to make of that
10:43:48 <Rolf1> jsled: From the changelog
10:43:57 <Rolf1> Where it really belongs
10:44:02 <Rolf1> for packaging purposes
10:44:18 <jsled> At what point is the version extracted and put in the changelog?
10:44:19 <Rolf1> you can even instruct the debian build tools to fetch that particular version
10:44:20 * andi5 remembers FUD
10:44:52 <Rolf1> jsled: When the packager writes the packaging defintion
10:47:36 <warlord> k
10:49:08 <warlord> thanks jsled
10:49:26 <andi5> cedayiv: i will include as many of the currently marked BPs as possible, starting to look for severity after 2.2.7
10:49:44 <andi5> is that ok?
10:50:11 <cedayiv> That's OK with me.
10:50:19 <andi5> Rolf1: how do those debian packages propagate the version numbers to the actual source code?
10:53:05 <Rolf1> #debian-mentor on oftc or #ubuntu-motu on freenode would be the right places to ask
10:53:29 <Rolf1> I'm usually the kind of guy that is glad not to have to understand every single inner working of a tool
10:53:44 <Rolf1> gnucash regularly makes me "unglad" here
10:54:05 <Rolf1> I know way too much of inner dealings of gnucash I never really cared about
10:54:54 <warlord> Well, if you just used the tarballs you'd never have trouble.
10:55:20 <warlord> (well, modulo bugs in the code or "missing features")
10:56:29 *** cedayiv is now known as cedayiv-afk
10:56:29 <Rolf1> exactly
10:56:40 <Rolf1> why do you think I put up with this?
10:57:16 <andi5> actually, i can second that... gnucash trunk builds on a variety of platforms and for a lot of people if special needs, but you need to dig a bit more than when compiling from a tarball
10:57:34 <Rolf1> I can compile from svn just fine
10:57:52 <andi5> honestly, what is the problem then?
10:58:04 <Rolf1> read up about pbuilder and packaging
10:58:11 <Rolf1> and why people prefer packaged software
10:58:38 <Rolf1> There is a reason for these tools
10:58:42 <andi5> who is people? ... do you mean yourself, or do you intend to distribute these packages?
10:58:46 <warlord> ./configure --prefix=/opt/gnucash works just fine.
10:59:05 <Rolf1> warlord in "warlord-world" mode again
10:59:21 <Rolf1> anybody using things differently than he is must be stupid or doing things incorreclty
10:59:25 <Rolf1> it *kills* me
10:59:31 <andi5> oh man, cannot you talk like normal people?
10:59:45 <andi5> you, meaning all three of us :)
11:00:17 <warlord> I didn't say you were stupid or doing things incorrectly, just pointing out that you don't need to use a package to get some of the features of a package when doing s/w development.
11:00:26 <Rolf1> It would be a start to acknowledge that nobody (that includes me) is stupid or doing something that they obviously should not attempt
11:00:44 <warlord> Oh, there are TONS of stupid people out there.
11:00:46 <Rolf1> I concede that generally you are more knowledgeable with gnucash and programming languages
11:00:58 <Rolf1> but I think I have more knowledge when it comes to packaging
11:01:02 <warlord> (I doubt you are one of them)
11:01:08 <andi5> mine is longer.... btw
11:01:24 <warlord> You certainly have more knowledge when it comes to debian/ubuntu packaging.
11:01:39 <warlord> I would bet I have more knowledge about RPM
11:01:49 <Rolf1> and generally, what makes me mad about gnucash is that everytime I touch it, I quickly find something that behaves strangely and when digging deeper I get into an ever deeper mess. That is unique to gnucash for me
11:02:20 <warlord> Oh, I've found strangeness in almost any package I try to build from source..
11:02:40 <Rolf1> There's weirdness you can shrug off and weirdness that makes things impossible to use
11:02:57 <Rolf1> gnucash does not work in the context I want to use it in for over a year
11:03:07 <Rolf1> despite great effort and some improvement
11:03:17 <andi5> actually, i am quite happy about gnucash, it build without a single modification on a quite obscure linux distribution.... it has never been tested on before :)
11:03:36 <andi5> that indicates that it cannot be _that_ broken
11:03:38 <warlord> Indeed, gnucash is quite portable
11:03:39 <Rolf1> If gnucash was usable I could use upstream packages and care less about helping with bleeding edge
11:04:01 <Rolf1> warlord: try cross-compilation
11:04:08 <Rolf1> I attempted that
11:04:14 *** nomeata has joined #gnucash
11:04:17 <warlord> why?
11:04:36 <Rolf1> gnucash stores information, right?
11:04:40 <warlord> The only time I've ever cross-compiled anything was building for a tuxscreen
11:05:03 <Rolf1> portable and luggable => you got something there
11:05:25 <Rolf1> gnucash or something similar on a PDA would make a lot of sense
11:05:27 <warlord> anyways, I have real work to do.
11:05:42 <Rolf1> but it's not only build time problems that will prevent gnucash from being used there
11:05:42 <warlord> I'll respond if I see anything worthwhile to respond to.
11:05:43 <andi5> what is luggable?
11:05:45 <warlord> Have a great day.
11:05:49 <Rolf1> andi5: portable
11:05:59 <Rolf1> warlord: bye
11:06:05 <Rolf1> and thanks for the support
11:06:15 <andi5> Rolf1: may i ask you one thing?
11:06:31 <Rolf1> go ahead
11:06:45 <Rolf1> but you've already used up "one thing" ;-)
11:06:56 <andi5> If you have learnt that discussing with gnucash devs is not worth the time, why do you always come back and tell us?
11:07:18 <Rolf1> good question ;-)
11:07:28 <andi5> i mean, i do not think so... but if that is your opinion, it will be hard for us to tell you different
11:07:33 <Rolf1> The reason is that I *need* something like it
11:07:36 <Rolf1> and badly
11:07:41 <Rolf1> a replacement takes time
11:07:49 <Rolf1> But I *am* working on it
11:07:56 *** esperegu has quit IRC
11:08:11 <Rolf1> I am pursuing several options
11:08:30 <Rolf1> If I had an alternative to the pain gnucash causes me, I'd dump gnucash in a minute
11:08:37 <Rolf1> Satisfying answer?
11:08:41 <Rolf1> Makes sense?
11:08:52 <warlord> Rolf1: I'll note that I do think most of your submitted patches have been committed.
11:08:53 <andi5> it does not answer my question at all
11:09:18 <Rolf1> warlord: That line gets boring after a while
11:09:31 <Rolf1> Ask jsled about patching and restarting from scratch
11:09:38 <Rolf1> s/and/or
11:09:55 <Rolf1> andi5: please clarify
11:10:17 <Rolf1> what kind of answer are you looking for or was your question rhetorical?
11:10:38 <andi5> Rolf1: to explain it: i suppose there are other ways to cope with buggy software, besides joining an irc channel, arguing, and bashing... especially when you think it is not worthwhile
11:11:09 <andi5> it may sound harsh, but when you want improvements, the best way still is to send patches
11:11:43 <andi5> or discuss the real problems... :)
11:12:00 <Rolf1> understood
11:12:07 <Rolf1> which is what I am trying to do
11:12:44 <andi5> ok, i hope it works for you =)
11:13:30 <Rolf1> not sure if discussion of real problems is possible with "gnucash" mindset. Sorry, I know this sounds harsh. I mean it as an honest food for thought. Which is one of the reasons I sometimes come back, both for help and because I hope that the message my eventually not be completely lost.
11:13:39 <Rolf1> I understand that hope is probably futile
11:14:05 <Rolf1> I can describe the problems I have
11:14:13 <Rolf1> See them brushed aside
11:14:38 <Rolf1> or in the best case "patches welcome, wfm, don't care"
11:15:20 <Rolf1> Since there are too many potholes for me, it makes gnucash difficult to impossible. Thus frustrating, sorry for bitching.
11:17:20 <andi5> out of curiosity, what about the other software packages you have tried? do they look promising but alpha, or dead or... just give me your opinion, if you want
11:19:42 <Rolf1> That would take very long
11:19:55 <Rolf1> I have tested probably two dozen or more packages
11:20:46 <Rolf1> basically, gnucash is still the program with the biggest following
11:21:03 <Rolf1> usually, the rule is "go with the big projects"
11:21:34 <Rolf1> german accounting is not funny and sexy for a programmer
11:21:43 <andi5> that is true :)
11:21:46 <Rolf1> which is why nothing completely suitable exists
11:21:58 <Rolf1> except for a few web-based solutions (even FOSS)
11:22:18 <Rolf1> But I'd never use a completely browser-based interface
11:22:19 <warlord> For the record, I do wish I could be given about 2 man-years of my time (paid) so I could effectively re-write most of gnucash.
11:22:45 <andi5> warlord: tell me when you start it
11:22:57 <Rolf1> Well, the word was always that money was not an issue
11:22:58 <andi5> maybe 23 months will be enough then ;-)
11:22:58 <warlord> After I win the lottery. ;)
11:23:04 <Rolf1> and bounties not needed
11:23:24 <andi5> do you know how expensive 2 years of warlord will be?
11:24:02 <andi5> i mean, bye me, i am cheap but useless :)
11:24:05 <andi5> buy
11:24:08 <Rolf1> FOSS is about making small steps incrementally
11:24:15 <Rolf1> rome wasn't built in a day
11:24:32 <Rolf1> andi5: actually, I had thought about sponsoring you
11:24:46 <andi5> oh?
11:24:55 <Rolf1> why not
11:24:57 <Rolf1> ?
11:25:22 <Rolf1> But I have really withdrawn the idea
11:25:24 <andi5> what did you say about german accounting? ;-)
11:25:33 <andi5> yes, i have this gnucash mentality, it is evil :)
11:25:39 <warlord> Rolf1: While rome wasn't built in a day there are times when you need to just raze a full city block and restart from scratch.
11:26:02 <Rolf1> It is hard for me to judge, but I have doubts if gnucash can be or should the the shoulders of the giant I want to stand on
11:26:04 <warlord> ... in particular when the underlying architectural choices are... problematic.
11:26:28 <Rolf1> even a complete rewrite starts with something small
11:26:33 <andi5> warlord: are they solvable with middle-sized steps?
11:26:51 <warlord> andi5: I honestly don't know.
11:26:52 <Rolf1> another good saying from Asia: a journey of a thousand miles starts with a first step
11:26:56 <warlord> I haven't thought about it.
11:27:11 <andi5> yes, but the question remains where to start
11:27:18 <andi5> and what route to take
11:27:36 <Rolf1> andi5: Let me show you something
11:28:21 <Rolf1> Can you access https://clients.leggewie.biz/oss/fibu/
11:28:23 <Rolf1> ?
11:28:34 * andi5 adds an ssl exeption :)
11:28:45 <andi5> yes
11:29:03 <Rolf1> Those are some ideas about a modular structure that some of us came up with
11:29:15 <Rolf1> It is written for us, but maybe you understand
11:29:31 <Rolf1> What you see is basically just the data model
11:29:36 <Rolf1> and just the core
11:29:48 <Rolf1> we argued very long for what should be in the core and what not
11:30:17 <Rolf1> if possible, the aim was to have functionality in modules whenever possible for reasons of maintainability
11:31:01 <warlord> It's very hard to modular SQL tables.. In particular when you have some module, X, that wants to add ancillary data to objects in another module.
11:31:55 <warlord> E.g., when you want to link an invoice to a transaction -- you could create a new linking table or you could add a new column to the txn. The latter is closer to accurate, but the former is more "sql-like"
11:32:25 <warlord> (and yes, deciding what is "core" is a very hard conversation)
11:33:14 <Rolf1> the linking table will be implicit
11:33:20 <Rolf1> in the generic grouping
11:33:33 <Rolf1> invoice itself will be one or even more modules
11:34:12 <Rolf1> core for us is tx, splits and accounts only
11:34:20 <andi5> please, excuse me, my time is running up and i want to get some things done, still.... have fun
11:48:21 *** esperegu has joined #gnucash
11:58:35 *** andi5 has quit IRC
12:01:19 *** andrewsw has quit IRC
12:10:08 *** donfede has joined #gnucash
12:21:23 *** aindilis has joined #gnucash
12:22:10 *** twunder has quit IRC
12:29:45 *** ErKa has quit IRC
12:42:36 *** ErKa has joined #gnucash
13:25:25 *** keyline has joined #gnucash
13:34:17 *** kielein has quit IRC
14:05:42 *** |gunni| has joined #gnucash
14:52:05 *** sjc has joined #gnucash
15:21:05 *** twunder has joined #gnucash
15:30:04 *** esperegu has quit IRC
15:35:40 *** nomeata has quit IRC
16:13:02 *** cedayiv-afk is now known as cedayiv
16:27:09 *** twunder has quit IRC
17:34:02 *** warlord is now known as warlord-afk
17:44:00 *** keyline has quit IRC
17:53:55 <todd> any unix people with gnucash 2.2.6 wanna tell me of ldd gnucash-bin shows libsp.so.?.?
17:54:18 *** |gunni| has quit IRC
18:04:13 *** twunder has joined #gnucash
18:14:56 *** twunder has quit IRC
19:28:18 *** ErKa has quit IRC
19:46:40 *** cedayiv is now known as cedayiv-afk
19:58:21 *** cedayiv-afk is now known as cedayiv
20:03:41 *** greg has joined #gnucash
20:05:46 *** greg has quit IRC
20:16:42 *** twunder has joined #gnucash
20:31:26 *** JimRaehl has joined #gnucash
20:42:32 *** jcreigh has joined #gnucash
20:43:20 *** cedayiv is now known as cedayiv-afk
21:09:30 *** paulproteus_ has quit IRC
21:14:25 *** mishehu has quit IRC
21:14:42 *** paulproteus_ has joined #gnucash
21:15:16 *** ben_goodger__ has quit IRC
21:19:04 *** sjc has quit IRC
21:22:01 *** ben_goodger__ has joined #gnucash
21:47:56 *** twunder has quit IRC
22:03:14 *** aindilis has quit IRC
22:05:03 *** aindilis has joined #gnucash
22:24:10 *** jcreigh has quit IRC
22:28:11 *** cedayiv-afk is now known as cedayiv
22:49:17 *** cedayiv is now known as cedayiv-afk
22:59:38 *** cedayiv-afk is now known as cedayiv
23:04:26 *** mishehu has joined #gnucash
23:22:11 *** paulproteus_ has quit IRC
23:50:27 *** cedayiv has quit IRC