2011-03-22 GnuCash IRC logs

00:59:28 *** slidesinger has quit IRC
01:47:05 *** zuhao has joined #gnucash
02:21:01 *** umbridge has joined #gnucash
02:22:33 <umbridge> hi all, we are looking for remote assistance in order to finalize a dedicated use of gnucash for our consultancy firlm, who could help ?
02:29:23 *** felle has quit IRC
02:38:17 *** esperegu has joined #gnucash
03:03:05 *** umbridge has quit IRC
03:32:35 *** ironwolf1 has quit IRC
03:40:40 *** ironwolf has joined #gnucash
03:57:37 *** ErKa has quit IRC
04:06:30 *** nodje has joined #gnucash
04:17:44 *** neil_ has joined #gnucash
04:29:07 *** bentob0x has joined #gnucash
05:00:09 *** gjanssens has joined #gnucash
05:00:09 *** gncbot sets mode: +o gjanssens
05:05:08 *** jay_t has joined #gnucash
05:51:31 *** kpreid has joined #gnucash
06:09:10 *** neil_ has quit IRC
06:09:59 *** esperegu has quit IRC
06:15:17 *** bentob0x has quit IRC
06:25:54 *** esperegu has joined #gnucash
06:56:03 *** kpreid has quit IRC
07:07:50 *** jay_t has quit IRC
07:28:07 *** kpreid has joined #gnucash
07:28:49 *** nodje has quit IRC
07:33:02 *** Askarii has joined #gnucash
07:35:10 <rand6> hi, is this the right forum for queries regarding GSoC?
07:46:12 <gjanssens> hi rand6, no this is the GnuCash irc channel.
07:46:25 <gjanssens> You should preferably use the gnucash-devel mailing list
07:46:48 <gjanssens> Most of the GnuCash people involved in GSoC are not on irc
07:47:03 <rand6> Alright thanks
07:47:12 <gjanssens> http://wiki.gnucash.org/wiki/Mailing_Lists
07:47:20 <gjanssens> You're welcome
07:55:44 *** Jozua_Judah has joined #gnucash
07:56:33 <Jozua_Judah> how do you deal with deposits on a vendor bill?
07:58:16 *** Askarii has quit IRC
07:58:40 *** warlord-afk is now known as warlord
07:58:44 <warlord> Jozua_Judah: Process Payment
07:59:12 <warlord> rand6: although you're certainly welcome to hang around here.
07:59:15 *** fbond has joined #gnucash
08:00:26 <Jozua_Judah> Warlord:Process Payment?
08:00:46 <Jozua_Judah> warlord: Just handle it like a normal billing item?
08:04:37 <warlord> Well, I assume that by "deposit" you mean that you've paid them money towards something. So yes, that would just be a "Process Payment", just like any other payment to the vendor would be.
08:05:09 <Jozua_Judah> I mean money I get back
08:05:24 <warlord> OH!
08:05:36 <warlord> That would be a credit note.. GnuCash doesn't have a good way to handle that..
08:05:46 <Jozua_Judah> I pay a breakage deposit of R2000 for accommodation at a camp site
08:06:03 <Jozua_Judah> if I don't break anything I get the money back
08:06:52 *** Jimraehl has left #gnucash
08:07:02 *** ceen has joined #gnucash
08:07:10 <warlord> AH. I'm afraid there's not a good way to handle that through the business features. But you can track it as a normal transaction. Bank -> Liabilities:Breakage Deposit. And then the reverse when you get your refund.
08:07:25 <warlord> Where in ZA are you, Jozua_Judah?
08:07:48 <Jozua_Judah> Stellenbosch
08:07:54 <Jozua_Judah> near Cape Town
08:07:59 <Jozua_Judah> cool thought so
08:08:01 <warlord> I know exactly where that is..
08:08:04 <warlord> It's a beautiful area.
08:08:11 <Jozua_Judah> no problems thanks
08:08:14 <Jozua_Judah> yip very
08:08:15 <warlord> (and great wines)
08:08:24 <Jozua_Judah> yes absolutely
08:08:26 <Jozua_Judah> :)
08:08:38 <warlord> It's about an hour's drive from Bantry Bay;.
08:09:27 <Jozua_Judah> yes
08:09:29 <Jozua_Judah> it is
08:09:33 <Jozua_Judah> g2g
08:09:45 <Jozua_Judah> thanks for the help again :)
08:09:50 <warlord> Later.
08:09:53 <Jozua_Judah> later
08:09:57 *** Jozua_Judah has quit IRC
08:31:47 *** Jimraehl has joined #gnucash
08:31:57 *** kpreid has quit IRC
08:40:34 *** gncbot` has joined #gnucash
08:40:46 *** Askarii has joined #gnucash
08:44:13 *** gncbot has quit IRC
08:52:20 *** warlord sets mode: +o gncbot`
08:52:23 <warlord> @nick gncbot
08:52:24 *** gncbot` is now known as gncbot
09:12:18 *** felle is now known as fell
09:14:26 *** Jimraehl has quit IRC
09:23:56 *** slidesinger has joined #gnucash
09:30:12 *** Jimraehl has joined #gnucash
09:34:42 *** Askarii1 has joined #gnucash
09:41:51 *** Askarii has quit IRC
10:01:48 *** Askarii1 has quit IRC
10:03:05 *** Askarii has joined #gnucash
10:06:26 *** Askarii1 has joined #gnucash
10:08:56 *** Askarii1 has quit IRC
10:10:08 *** Askarii1 has joined #gnucash
10:13:26 *** Askarii has quit IRC
10:17:25 *** Askarii has joined #gnucash
10:24:16 *** Askarii1 has quit IRC
11:07:06 *** Francisco has joined #gnucash
11:11:46 *** bentob0x has joined #gnucash
11:15:30 *** mr-rich has quit IRC
11:26:34 *** Askarii has quit IRC
11:35:40 *** mr-rich has joined #gnucash
11:53:37 <mr-rich> Question: can GnuCash interface with any POS software?
11:56:51 <warlord> Only if said POS software can export QIF or OFX, in which case you can import it into gnucash.
12:03:48 *** Askarii has joined #gnucash
12:06:05 *** zuhao has quit IRC
12:06:18 *** zuhao has joined #gnucash
12:13:53 *** fell is now known as fell_afk
12:20:27 *** slidesinger has quit IRC
12:27:18 <Antisoche> I'm getting an error doing a 'make install' from src/html - .libs/libgncmode-html.lai is not found. Any ideas? There's a .so and a .soT ...
12:28:02 <warlord> Antisoche: what OS/Distro?
12:28:48 <warlord> did you do a 'make' first?
12:28:56 <Antisoche> This is under Kubuntu natty. I think the initial fault is here: "libtool: link: /usr/lib/libgmodule-2.0.la' - No such file or directory
12:29:26 <warlord> Have you updated anything since your last full (successful) make & install?
12:29:42 <Antisoche> Yeah, I just updated. :(
12:30:22 <Antisoche> Any idea what libgmodule is?
12:30:29 <warlord> it's a glib thing.
12:30:31 <Antisoche> The error actually comes from 'sed'
12:30:47 <warlord> you'll need to "make clean", reconfigure, and rebuild gnucash
12:31:14 <Antisoche> I did that to clear up another error...
12:31:35 <Antisoche> Looks like libglib2.0-dev may have been broken ...
12:31:51 <warlord> could be
12:32:58 <Antisoche> Hmm. It moved to /usr/lib/i386-linux-gnu. Would 'configure' pick that up you think?
12:34:17 <warlord> it should, yes
12:34:38 <warlord> (assuming it doesn't re-use the configure cache)
12:34:59 <Antisoche> I did a 'sudo make clean' ... how do I start; autoconf ? autoupdate ?
12:35:13 <Antisoche> autoscan ?
12:35:45 <warlord> why did you do that? Why are you building as root?
12:36:09 <Antisoche> I make install as root; that tends to create root-owned files in the tree
12:36:29 <warlord> Oh.
12:36:43 <warlord> See, if you used opt-style you wouldnt have to install as root ;)
12:37:14 <Antisoche> I install into /usr/local; same diff. I just compile so infrequently I haven't changed permissions.
12:38:15 <Antisoche> autogen.sh is missing and I can't 'svn up' it?
12:38:47 *** esperegu has quit IRC
12:39:07 <Antisoche> nevermind ... I was in the src/ directory
12:39:09 <warlord> ????
12:42:44 <warlord> ok
12:48:30 *** esperegu has joined #gnucash
12:50:21 <Antisoche> I had a bunch of root owned files in various flavors. Removed all of those and in the middle of a compile now.
12:53:28 <Antisoche> Same problem - sed: can't read /usr/lib/libgmodule-2.0.la
12:53:40 <Antisoche> Time for a symlink I think...
12:55:15 <warlord> Maybe. Which .la file is referencing that?
12:56:26 <Antisoche> Actually, I probably need to re-rerun autogen, but for now let's try this. Error is:
12:57:03 <Antisoche> libtool: link: gcc -shared .libs/swig-report-gnome.o [...] -o .libs/libgncmod-report-gnome.so
12:57:23 <Antisoche> /bin/sed: can't real /use/lib/libglib-2.0.la: No such file or directory
12:59:12 <Antisoche> What is the configure cache file, config.status ?
13:01:07 <warlord> no, it would be something like auto4mate.something
13:01:10 <warlord> (it's a directory)
13:01:39 <warlord> Hmm, I wonder what .la is referencing /use/lib/libglib-2.0.la ?
13:02:53 <Antisoche> Ok. Well, I did a 'sudo make clean', 'find . -uid 0' (nil), './autogen.sh', and now 'LINGUAS=C ./configure --enable-ofx --enable-aqbanking --disable-static --enable-debug --enable-dbi --no-create --no-recursion --enable-compile-warnings'
13:03:30 <Antisoche> I think there may be a problem with the webkit renderer, so I want to use gtkhtml. Most of the other settings are what Ubuntu is building with.
13:03:59 <Antisoche> Running 'make' now...
13:05:06 <Antisoche> How do I tell what '.la' is referencing those files? It's in the reports/ directory somewhere ...
13:05:18 <warlord> If it still fails, then: find /usr -name \*.la | xargs grep libglib-2.0.la
13:05:52 <Antisoche> .la files are what? It looks like a shell script fragment
13:08:55 <warlord> .la == libtool archive
13:11:16 <Antisoche> Okay, getting the same error...
13:13:27 <warlord> Run that find command?
13:15:39 <Antisoche> bunches of them
13:18:04 <Antisoche> Stuff in /usr/lib/enchant ... /usr/lib/goffice ... /usr/lib/NetworkManage /usr/lib/orbit, and a lot of /usr/local/lib/gnucash files. Curiously, even /usr/lib/i386-linux-gnu/libgmodule-2.0.la does, but I hope that's a conditional
13:18:59 <warlord> Hmm... I wonder if it's pulling anything from /usr/local/lib/gnucash ?
13:19:14 <Antisoche> Er, that's not full a path, just the binary name. Are you looking for a hardcoded path?
13:19:29 <warlord> Yes
13:19:38 <warlord> But your path above was wrong?
13:20:20 <Antisoche> From the compile, yes. That was calling out /usr/lib when the binary is in /usr/lib/i386...
13:20:36 <Antisoche> So whatever that libtool is calling
13:20:38 <warlord> No, I mean: <Antisoche> /bin/sed: can't real /use/lib/libglib-2.0.la: No such file or directory
13:20:39 <warlord> You said: /use/lib when you meant /usr/lib
13:20:54 <warlord> That's what I meant by "your path above was wrong"
13:21:07 <Antisoche> Ah. I'm compiling in a Virtual Machine and can't C&P - transcription error.
13:21:35 <Antisoche> "can't real" is also a typo, it said "can't read"
13:22:06 <warlord> heh
13:22:21 <warlord> You don't have real C&P into your VM? Weird. Vmware provides that ;)
13:23:00 <Antisoche> Kubuntu is shipping a very new x.org
13:25:04 <Antisoche> I have a VMWare license, but VirtualBox is easier and covers my needs, as much as my stupid CPU will allow. :(
13:25:20 <warlord> heh
13:26:54 <Antisoche> I've always run AMD processors (except my 286-20, which was actually made by Harris) and decided I'd get an Intel this time. 4-core HT, etc, etc, and the thing can't virtualize 64-bit OSes. The AMD I gave to my parents could do it. Bah!
13:27:19 <Antisoche> As in, my old AMD that I thought I was upgrading from ...
13:27:33 <warlord> LOL.
13:27:50 <warlord> Your Intel Chip doesn't have the VME instruction?
13:28:20 <Antisoche> It has the VME flag in /proc/cpuinfo
13:28:22 <jsled> what could you have possibly bought that doesn't have virt support?
13:28:35 <warlord> Is it only a 32-bit CPU?
13:29:05 <Antisoche> Core2 Quad Q8200. It's like the Q8100 and Q8300 and Q-everything else will do it, but the 8200 is somehow "special"
13:29:16 <warlord> WEIRD
13:29:36 <Antisoche> Not my initial reaction, but sure. It's a pricing thing from what I recall.
13:30:05 <jsled> heh
13:31:23 <Antisoche> http://ark.intel.com/Product.aspx?id=36547
13:31:37 <Antisoche> VT-x: No. Apparently, no HT either, so I was wrong about that.
13:31:49 <Antisoche> 4 threads, 4 cores, so it's 1:1
13:31:57 <Antisoche> I hate this thing more and more every day. :p
13:32:08 <warlord> LOL
13:32:15 <warlord> Time to replace the CPU?
13:32:28 <Antisoche> http://www.tomshardware.com/news/Intel-E5200-Q8200,6302.html
13:32:55 <Antisoche> I don't think it's worth it. Intel CPUs are overpriced. It's cheaper to buy another PC.
13:33:45 <warlord> I've been happy with my dual-cpu, quad-core Opteron system.
13:34:07 <warlord> (except that I can effectively hang the system for periods of time with lots of disk access)
13:34:08 <Antisoche> Yeah - I got tired of building PCs and had pretty good luck with the HP AMD box I got; just got lucky I guess.
13:34:56 <Antisoche> warlord: I'm having similar problems. Right now it's looking like collectd is my problem, but I really need to focus on one problem at a time here...
13:35:05 <warlord> heh
13:35:42 <Antisoche> So my compiling problem is coming from report/report-gnome, and badly, if I type 'make' the error gets bypassed.
13:36:27 <Antisoche> Hmm, it's not generating a 0 byte file ...
13:36:42 <warlord> what's wrong with report/report-gnome?
13:36:56 <Antisoche> That's where the 'sed' error is coming from
13:37:27 <Antisoche> Or 'libtool' error I guess it is.
13:38:00 <Antisoche> libtool: link: gcc -shared .libs/swig-report-gnome.o [...] -o .libs/libgncmod-report-gnome.so
13:38:21 <Antisoche> /bin/sed: can't read /usr/lib/libgmodule-2.0.la: No such file or directory
13:38:47 <Antisoche> libtool: link: /usr/lib/libgmodule-2.0.la is not a valid libtool archive
13:38:56 <Antisoche> ... and then make unwinds
13:39:18 <Antisoche> The .so is getting created; it's 300k or so
13:43:48 *** jay_t has joined #gnucash
13:44:48 *** jay_t_ has joined #gnucash
13:44:59 *** jay_t has left #gnucash
13:45:59 <Antisoche> Aha. I think I may have found the solution ...
13:46:09 <Antisoche> make || make || make
13:46:58 <Antisoche> 'stylesheets' and 'test' have the same problems but so far it looks like things are limited to the 'report/' directory.
13:47:56 *** neil_ has joined #gnucash
13:48:22 <warlord> You need to find which .la on your system is referencing /usr/lib/libgmodule-2.0.la
13:48:40 <warlord> (and said .la might be in /usr )
13:48:47 <Antisoche> er, and 'src/gnome'. And 'import-export/qif-import'. warlord: okay.
13:53:18 <Antisoche> Okay, so there's still quite a few. /usr/lib/{enchant,goffice,NetworkManager,orbit-2.0} and all the gnucahs ones in /usr/local/lib
13:56:14 <Antisoche> Who is calling 'sed' is what I want to know ...
13:57:02 <warlord> libtool.
13:57:17 <warlord> The bug is that enchange, goffice, and orbit -dev packages are all broken.
13:57:34 <warlord> goffice and orbit are the most important ones.
13:57:43 <warlord> In think you may just need to create a symlink.
14:01:37 <Antisoche> Looks like I 'make'd past it. Doing an install now. We'll see what happens next ... :-\
14:02:17 <Antisoche> No Such File errors. yay
14:04:51 <warlord> cool
14:09:30 <Antisoche> I don't really understand why it gets all the way to the reporting before it bails.
14:10:30 * Antisoche dislikes auto*
14:11:07 <Antisoche> I removed the output file from .libs and did a make again, now I get a link error elsewhere. Isn't the whole point of make to build stuff that's not there?!
14:12:29 <warlord> Well, some of it depends on autogen and configure's results
14:13:38 <Antisoche> All I know is that when I write a Makefile, if I remove a .so and run make, it gets built...
14:14:19 <warlord> Yes, and if you remove the .la it will rebuild too.. But removing the .libs dir is sorta like removing the Makefile's include-dir
14:14:35 * Antisoche sighs. So much complexity...
14:14:59 <Antisoche> I didn't remove the .libs dir. I removed the .so that it built and put in .libs before giving me the libtool/sed error.
14:15:22 <warlord> Why would you do that?
14:15:50 <Antisoche> Because if that file exists, when I type 'make' again, whatever that 'sed' is supposed to do doesn't get done.
14:16:27 <Antisoche> And actually, it appears it doesn't matter if that file even exists. If it's missing, it doesn't get rebuilt.
14:16:51 <Antisoche> It just thinks everything's ok. Like it set a flag in .deps and just moved on.
14:16:59 <warlord> Yeah, the dependency is based on the .la file, not the .so
14:18:47 <Antisoche> Ah. So I must have a 0-byte .la file somewhere.
14:19:57 <warlord> Maybe.
14:20:03 <warlord> Try "make clean" and start over?
14:20:47 <Antisoche> Shouldn't there be a "%(.la): %(.la=.so)" type rule somewhere (not the right syntax, but conceptually...)
14:21:34 *** zuhao has left #gnucash
14:21:54 <Antisoche> So, anyway, whatever Ubuntu is up to with moving that stuff around I don't know, but the symlinks got me going again. Thanks for listening.
14:22:27 <Antisoche> "$(LA): $(LA:%.la=%.so)" looks closer...
14:23:11 <warlord> Well, part of the magic happens in libtool
14:24:39 *** bentob0x has quit IRC
14:50:47 *** fell has joined #gnucash
14:50:48 *** gncbot sets mode: +o fell
14:50:48 *** ktdreyer has quit IRC
14:58:46 *** jay_t_ has quit IRC
14:59:01 *** fell_afk has quit IRC
15:08:51 <gjanssens> warlord: I think I have a fairly simple solution ready for the current build system issue
15:08:57 <gjanssens> I have updated some build scripts
15:09:04 <warlord> Okay..
15:09:19 <gjanssens> to make sure that each run actually uses mingw as defined by $MINGW_DIR
15:09:50 <gjanssens> On the server, we should now duplicate the c:\soft directory
15:10:23 <gjanssens> and configure one to build based on trunk and one to build based on 2.4
15:10:42 <warlord> I'm happy to make a c:\soft-2.4
15:10:47 <gjanssens> both will have a packaging directory
15:10:55 <gjanssens> that's a good name
15:11:10 <gjanssens> but both will have a slightly different custom.sh
15:11:23 <gjanssens> for trunk, it can stay as it is now
15:11:34 <warlord> Well, which is the "master" packaging directory? I.e., which is the one that the system calls for the "daily_build.bat"?
15:12:04 <gjanssens> for soft-2.4 it will need to override the svn branch and the global dir
15:12:34 <gjanssens> daily_build.bat should use what we have now, meaning it should build trunk
15:13:30 <gjanssens> if we want to add a daily/weekly build for 2.4 as well we can add this in daily_build.bat (for daily builds) or create a weekly_build.bat (for weekly builds)
15:14:12 <gjanssens> that bat file should just cd into the appropriate packaging directory an run daily_build.sh
15:14:23 <gjanssens> The rest will be taken care of by the modified scripts.
15:14:56 <gjanssens> I'm just considering adding a third base directory for the tag builds
15:15:05 <warlord> And what about the tags? How does a tag-build know whether to build from c:\soft or c:\soft-2.4 ?
15:16:02 <gjanssens> Perhaps even set this up completely programmatically
15:16:21 <gjanssens> Have it check out the proper packaging directory for the tag to be built
15:16:54 <gjanssens> Add a custom.sh for tags (setting the proper global dir for example)
15:17:08 <gjanssens> And then run install.sh from there.
15:18:03 <gjanssens> Since tags import everything anyway, they should also install their own mingw environment (I hope), so it should always bootstrap itself properly.
15:18:19 <gjanssens> I'll dig into the tags story a bit further
15:18:24 <gjanssens> just to be sure
15:19:22 <warlord> Basically, it would be nice if build_package.sh could do all the magic.
15:23:14 <gjanssens> I think the build_tags.sh script can be easily adapted for the above scheme.
15:24:19 <gjanssens> I notice that the DOWNLOADS_DIR is copied for each tag. Why is that done instead of just reusing the downloads dir (ie configure it in custom.sh to point to the global downloads dir ?)
15:25:39 *** KaiForce has joined #gnucash
15:25:39 <gjanssens> That's not really relevant to the problem at hand though, but I was just wondering
15:29:00 <warlord> Because I didn't want a tag-build to affect the main build.. So it was easier to do it that way.
15:30:33 <gjanssens> Ok.
15:31:41 <gjanssens> It seems the tag builds already do what I said above, they only use another path than I proposed.
15:31:52 <warlord> okay.
15:31:59 <gjanssens> So the tag builds script should be run from c:\soft.
15:32:27 <gjanssens> It will create the necessary packaging directory based on the tag and start the install from there.
15:33:18 <gjanssens> So do you want to build daily or weekly 2.4 builds ?
15:33:49 <gjanssens> In that case we should add something to daily_build.bat or add a weekly_build.bat on the same principle
15:34:23 <gjanssens> Otherwise, it's now just a matter of cleaning out the c:\soft directory so it picks up the new gcc
15:34:48 <warlord> Will the tag-build install gcc?
15:34:58 <gjanssens> Yes
15:35:07 <gjanssens> It really starts from scratch
15:35:07 <warlord> Really?!? Sweet.
15:35:55 <warlord> I don't think we need a daily 2.4 build. We could do a weekly one. Not sure if that's easily done off the existing bat-file.
15:36:11 <warlord> I suppose I could configure for a second (weekly) bat file.
15:36:23 <gjanssens> Sure, I'll create you a weekly_build.bat file
15:36:26 <warlord> okay.
15:36:38 <gjanssens> Can you setup a soft-2.4 directory ?
15:38:19 <gjanssens> As far as I'm concerned, you can start by copying c:\soft
15:38:44 <gjanssens> After that, we need to modify it's custom.sh
15:39:01 <gjanssens> and afterwards probably run reset.sh
15:39:22 <gjanssens> Does that remove the repos directory also ? If not, we should remove it manually
15:39:50 *** neil__ has joined #gnucash
15:44:05 *** sjc has joined #gnucash
15:45:04 <gjanssens> The script changes have been commited in r20452
15:45:27 *** trnto_dev has joined #gnucash
15:46:26 <warlord> I dont recall if reset.sh removes repos.
15:46:42 *** trnto_dev has quit IRC
15:46:45 <gjanssens> Hmm, this may have been a candidate for backporting but I forgot to add BP...
15:47:44 <gjanssens> Anyways, the custom.sh in c:\soft\packaging would need these additions:
15:48:04 <gjanssens> REPOS_URL="http://svn.gnucash.org/repo/gnucash/branches-2.4"
15:48:14 *** neil_ has quit IRC
15:48:26 <gjanssens> GLOBAL_DIR=C:\\soft-2.4
15:48:53 <gjanssens> And that's probably it, though we may want to reuse HH from c:\soft
15:50:19 <gjanssens> As I'm at it, I may as well try to setup c:\soft-2.4 myself if you haven't started yet. That way I can check if I can still login...
15:51:33 <warlord> I haven't done it yet, no.
15:51:35 <warlord> So go ahead.
15:51:58 <gjanssens> Ok
15:59:35 <gjanssens> There's one important caveat though: we should make sure that only one build is running at the time
15:59:45 <gjanssens> Or at least only builds that use the same mingw
16:00:22 <gjanssens> How is Windows dealing with daily and weekly "cron" jobs ?
16:00:34 <gjanssens> Are they started in parallel or sequentially ?
16:00:43 <gjanssens> The former will give problems.
16:00:58 <warlord> Right now there is no weekly; there is only daily.
16:01:10 <gjanssens> I know, but I'm anticipating
16:01:12 <warlord> (which is why it would be nice to kick the 'weekly' off the daily.bat if we can)
16:01:52 <gjanssens> Nice indeed, but my windows cmd fu is rather weak.
16:02:07 <gjanssens> Do you happen to know how to test for the day of the week ?
16:02:30 <warlord> Well, we can kick off a 'weekly_build.sh' that will only allow itself to be run on e.g. Sunday.
16:02:47 <gjanssens> You were a split second faster than I :)
16:03:11 <warlord> E.g. date +%u
16:03:32 <warlord> if [ `date +%u` != 6 ] ; then exit ; fi
16:09:28 *** markjenkinsparit has joined #gnucash
16:12:31 *** neil__ has quit IRC
16:13:34 *** neil_ has joined #gnucash
16:13:35 <gjanssens> warlord, it should probably 7 if you want it to run on Sunday %u is [1..7] with 1 being Monday
16:14:48 <warlord> I was just proposing Sunday as an option. Perhaps we want it on Monday (considering it gets run at 2am?) I'm not sure.
16:19:39 <gjanssens> In the end it doesn't really matter, but I like the idea of building at the end of the weekend, so Monday is good for me.
16:21:33 *** Francisco has left #gnucash
16:21:35 <warlord> Yeah, I think 2am Monday would be best, to let people get stuff patched on Sundays.
16:38:59 <gjanssens> Monday it is.
16:39:14 <gjanssens> The build server is still copying...
16:39:35 <gjanssens> There's one more little thing to keep in mind.
16:39:59 <gjanssens> The build scripts upload the installer and build log to the webserver
16:41:13 <warlord> Yep. We will need to think about how to better name the builds.
16:41:27 <warlord> Maybe by adding a branch-name if it's not "trunk"?
16:41:52 <gjanssens> Or uploading to different paths
16:42:12 *** ErKa has joined #gnucash
16:42:21 <gjanssens> It all goes into the directory "trunk" now, but that doesn't really make sense for branch and tag builds does it ?
16:42:36 <warlord> Ah, true, there is that.. Currently it all uploads into win32/trunk. Perhaps I should upload into win32/trunk, win32/2.4, and win32/tags?
16:42:54 <gjanssens> That would be my proposal indeed
16:44:43 <warlord> I'll create the appropriate directories on the server just now.
16:45:18 <gjanssens> Ok, but I was thinking of creating the directories automatically just before uploading.
16:45:19 <warlord> Done
16:45:37 <gjanssens> Or is ssh to the server not working, only scp ?
16:46:32 <warlord> I dont know... I'd have to remember how it's limited...
16:47:17 <gjanssens> I will try manually from the build server before changing any scripts then
16:47:31 <warlord> Go ahead..
16:47:39 <warlord> Although I did already create the 2.4 and tags dirs.
16:48:17 <gjanssens> Yes, it's just for future branches, so we don't have to remember...
16:49:07 <warlord> yep
16:50:19 *** markjenkinsparit has left #gnucash
17:03:30 <gjanssens> Ok done.
17:03:42 *** KaiForce has quit IRC
17:03:46 <gjanssens> warlord: fyi, ssh access is blocked (good!)
17:04:08 <gjanssens> So I used a small hack via scp
17:04:25 <gjanssens> I created some test directories while figuring it out.
17:04:40 <gjanssens> Can you remove win32/2.5 (and subdirectories) for me again from the webserver ?
17:07:16 *** ErKa has quit IRC
17:08:43 *** ErKa has joined #gnucash
17:10:18 <warlord> gjanssens: I thought it would be..
17:10:20 <warlord> Sure.
17:11:07 <warlord> Hmm, should I rename "tags" to "releases"?
17:12:32 <gjanssens> Good idea. I'll add it to the script. If you just remove the tags directory, we can test at the next tag build if the script properly creates the directories.
17:14:20 <warlord> Okay.
17:14:42 <warlord> Done
17:15:14 <gjanssens> Thanks
17:16:22 <gjanssens> BTW, I intend to backport all these build script changes to 2.4. I fear we may hit subtle bugs if the branch is not aware of the potential mingw swapping going on on trunk.
17:16:34 <warlord> okay.
17:19:24 *** ErKa has quit IRC
17:19:41 *** ErKa has joined #gnucash
17:28:57 *** esperegu has quit IRC
17:35:13 *** ErKa has quit IRC
17:53:41 <gjanssens> warlord: I think everything should be in place now
17:54:04 <warlord> okay.
17:54:29 <warlord> Shall we wait overnight to test it out? Or do you want to manually fire off a trunk build just to test it?
17:54:30 <gjanssens> The daily_build.bat script has changed. Do you have to renew the Windows "cron" stuff, or will it pick up the new script automatically ?
17:54:50 <warlord> It'll pick up the new script a day later.
17:54:51 <gjanssens> I would just wait for the overnight build to see what happens
17:55:04 <warlord> unless you manually 'svn update' /c/soft/packaging
17:56:22 <warlord> the first thing daily_build.sh does is an svn update, but that doesn't help in terms of updating the daily_build.sh ;)
17:56:26 <gjanssens> I have run the svn update. I prefer not to loose one night.
17:56:51 <warlord> Okay.
17:57:02 <warlord> We'll find out for sure in about 9-12 hours
17:57:31 * Antisoche is back
17:57:59 *** sjc has quit IRC
18:05:37 *** sjc has joined #gnucash
18:12:58 <gjanssens> Ok, that's it for today. Night is calling here.
18:13:01 <gjanssens> See you tomorrow
18:13:42 *** gjanssens has quit IRC
18:28:57 *** rpg has joined #gnucash
18:31:15 <Antisoche> Are there any sort of perl-like hash primitives in the reports?
19:19:25 <warlord> Antisoche: I'm not sure. There are certainly vectors and records. You could certainly build a binary-tree to imitate a hash...
19:20:46 <Antisoche> I'm using something apparently called an 'alist'. We'll see how it goes...
19:21:36 <Antisoche> http://www.gnu.org/software/guile/manual/html_node/Association-Lists.html#Association-Lists
19:21:42 <warlord> Ah, yeah, alist might be close, too, although I suspect it's still a linear search.
19:22:20 <warlord> Antisoche: http://www.math.grin.edu/~stone/courses/scheme/readings/association-lists.xhtml for another example
19:25:15 <Antisoche> It looks like this may be working ...
19:25:50 <Antisoche> Of course, I have the wrong datatypes again...
19:33:18 <Antisoche> If you don't mind saving me some searching, given a '(xaccSplitGetAmount)' (which is a numerator/denominator thing), how do I get the monetary value of it?
19:37:24 <warlord> What do you mean? xaccSplitGetAmount returns you a gnc_numeric
19:42:56 <Antisoche> Right. I can't feed that into a (+ 1 ...) type structure. Maybe I need the gnc_numeric adder function.
19:43:41 <Antisoche> But at some point I want to display it as a $ or CMDTY type value...
19:43:44 *** rpg has quit IRC
19:44:15 <Antisoche> I can find it eventually... register.scm is doing something new with an 'accumulator' that's unfamiliar.
19:45:35 *** ErKa has joined #gnucash
19:45:55 <warlord> right, you should wrap the Amount into a monetary_collector and use the accumulators. It'll make your life easier, in general. And yes, you will need to use the gnc_numeric functions for +/-/*/รท
20:04:56 *** rpg has joined #gnucash
20:08:40 *** zuhao has joined #gnucash
20:29:36 *** zuhao has quit IRC
20:29:45 *** zuhao has joined #gnucash
20:49:44 *** ErKa has quit IRC
20:57:25 *** kpreid has joined #gnucash
21:02:55 *** sjc has quit IRC
21:03:55 *** Jimraehl has quit IRC
21:05:01 *** Jimraehl has joined #gnucash
21:06:24 <Antisoche> I think I disagree
21:07:32 <warlord> this seems to be a trend ;)
21:07:48 <Antisoche> There we go ... all It takes is for me to make an ass of myself and I figure it out. :)
21:07:59 <warlord> hahahahaha
21:08:01 <warlord> :-D
21:08:36 <Antisoche> scheme doesn't really let you parenthesize randomly like C does...
21:08:54 <warlord> that is correct
21:16:26 <Antisoche> If I put a (define var ...) in the global section of a script it persists across different script invocations? Reloads, same report in other tabs, etc?
21:20:14 <warlord> for the most part, yes.
21:20:31 <warlord> It all depends on the environment.
21:21:47 <Antisoche> Cool 404, btw: https://github.com/Gnucash/gnucash/commit/
21:30:30 *** rpg has quit IRC
21:30:40 <Antisoche> okay ... so now I want to iterate over a list. Is the idea that I pull out the cdr and recursively look at the car as long as the car type is a list?
21:31:53 <warlord> Or you can use for-each
21:31:56 <warlord> or map
21:32:09 <warlord> or .... there are lots of ways to iterate (or recurse) over a list.
21:36:53 *** ceen has quit IRC
21:37:49 <Antisoche> GnuCash is using Guile, right?
21:38:20 <Antisoche> The Guile docs only list 'do', 'while', and 'named let' as iteration mechanisms.
21:39:30 <warlord> yes, but
21:39:42 <warlord> guile> for-each
21:39:42 <warlord> #<primitive-generic for-each>
21:44:00 <Antisoche> Same as array-for-each listed under 'Array Procedures'?
21:45:44 <warlord> possibly.
21:47:19 <Antisoche> They use for-each in their Tortoise tutorial.
21:49:59 <Antisoche> Okay - clearly the guile docs expect some level of scheme familiarity. Moving on ...
22:04:00 <warlord> Antisoche: I recommend you google for "MIT Scheme Reference" which is a great resource.
22:07:03 *** warlord is now known as warlord-afk
22:07:22 <Antisoche> I looked for "scheme 'for-each' example" and got very little useful results.
22:07:36 <Antisoche> I think Google is trying to be too clever for its own good.
22:18:03 <fell> Oops, Internal Server Error @ bugzilla
22:20:05 <fell> Antisoche: guile ~ scheme ~ lisp
22:34:56 <Antisoche> Bugzilla has been doing that lately.
22:45:27 *** rpg has joined #gnucash
22:46:43 *** Askarii has quit IRC
23:04:15 <Antisoche> Can someone give me the nickel tour of the 'collector' concept?
23:04:59 <Antisoche> Specifically, register.scm line 513 has: (debit-amount 'add split-currency split-amount)
23:05:33 <Antisoche> Here, "(debit-amount" looks like a function name but it's the name of a passed variable.
23:05:41 <Antisoche> "debit-amount" appears nowhere else in this file.
23:06:18 <Antisoche> Oh, wait...
23:10:15 <Antisoche> Okay, I get that it's a function name, but I still don't see what it is. It's either 'debit-collector' or '(gnc:make-commodity-collector)' and I don't see what either of those are.
23:13:49 *** harleypig has joined #gnucash
23:15:03 <harleypig> Is there a way I can force the APPID and APPVER to be blank in the aqbanking user setup? If I leave them blank in the gui, it appears to default to QWIN 1200 and my bank is returning 'not supported' messages for that.
23:16:10 <Mer|in> general accounting question: when I return an item to the store, I put it as a reversing transaction against the expense account type? it's not considered an income?
23:18:17 <fell> harleypig: hbci or ofx user?
23:18:24 <harleypig> fell: ofx
23:20:38 <Antisoche> Mer|in: http://webcache.googleusercontent.com/search?q=cache:Y00-yCr1akYJ:www.gnucash.org/docs/v2.4/C/gnucash-guide.pdf+gnucash+documentation+wash+account&cd=3&hl=en&ct=clnk&gl=us&client=ubuntu&source=www.google.com#215
23:21:07 <fell> I /believe/ somewhere in http://wiki.gnucash.org/wiki/OFX_Direct_Connect_Bank_Settings or reated pages it is explained.
23:21:49 <Antisoche> Mer|in: It's in Chapter 16 - "Wash Account". Search for the GnuCash Tutorial and Concepts Guide. It's a PDF so not very link-friendly unfortunately.
23:22:44 <Mer|in> Antisoche: thanks
23:25:36 <fell> We have more: http://svn.gnucash.org/docs/
23:26:17 <Antisoche> harleypig: I think you may have to edit the config manually. I believe if you put a space it'll erase the old values, but I'm not sure if your bank will like a blank space any better.
23:26:25 <harleypig> fell: I already found that page, my credit union isn't there.
23:26:35 <fell> not only pdf, different languages, Versions...
23:26:57 <harleypig> Antisoche: the file in ~/.aqbanking/users ?
23:27:17 <Antisoche> Google doesn't like to index the GnuCash website it seems. It would be nice if it did.
23:27:46 <Antisoche> harleypig: I'm not sure. I fought with it for a while and then either figured out that my bank doesn't support it or I just gave up.
23:30:58 <fell> Harleypig: which apps in which versions are supported by your bank?
23:31:03 <Mer|in> Antisoche: makes sense - so that would be considered equity then?
23:32:24 <Antisoche> Mer|in: I don't think equity is right. If the store credited your card, for instance, you'll subtract from the expense and credit the card.
23:33:01 <Antisoche> Mer|in: I guess I didn't really read that thing I pointed you to. I just happened to fix a bug in the math recently so had it in mind. :)
23:33:01 <harleypig> Antisoche: I added MYAPPID and MYAPPVER then changed those values in the config file to char appId = "" and the ofx.log shows those values still being set to QWIN and 1200 ...
23:33:22 <harleypig> fell: I'm pretty sure Quicken 2009, but I can't find an appver for it.
23:34:57 <Antisoche> harleypig: I assumed the GUI was seeing the value as empty and not changing it. So if you set it to 'foo' you could never clear it. If you take it out of the config file I don't know where it would get the value. Make sure the program isn't running when you edit the config; otherwise, I don't know what's happening.
23:35:56 <harleypig> Yeah, I made sure gnucash was closed before editing that file. I changed it to char appId=" " and got a network connection error.
23:36:19 <Antisoche> harleypig: Well - if you're editing the file don't put in a blank space. :)
23:36:27 <harleypig> Antisoche: :] right
23:36:43 <Antisoche> Either make it empty or just remove it. Not sure what the prog will do w/no value tho.
23:37:31 <fell> I believe QWIN means Quicken for Windows; and then you must play with the Version.
23:37:46 <Antisoche> Either way, I gotta run. Good luck all and if warlord or anyone wants to educamate me about scheme collectors I'll be logging...
23:38:21 *** slidesinger has joined #gnucash
23:39:22 <harleypig> fell: I just tried QWIN 1800 and got "Intuit Quicken 2009 is not supported by Mountain America Credit Union" which is weird, because they just sent out a notice stating you had to upgrade to that ...
23:39:32 <harleypig> anyway, I've got to get home. Thanks for the help
23:40:27 <fell> You are welcome.
23:44:12 <fell> Mer|in: About your original questions I would answer both with yes. you neutralize the expense.
23:52:13 <Mer|in> fell: so then does it go into a single "wash" account or is there any GAAP-type reason to use a wash account that matches the expense type?
23:52:35 <Mer|in> e.g. for Expenses:Groceries have an Equity:Suspense:Groceries
23:54:23 <Mer|in> fell: nm i just realized you were referencing different questions
23:58:56 <fell> wash is ambiguous for me. Can you send me the link to the html docs?
23:59:59 *** rpg has quit IRC