2015-12-19 GnuCash IRC logs

00:15:15 *** thomasr has quit IRC
01:35:38 *** calp has quit IRC
01:36:44 *** calp has joined #gnucash
01:39:59 *** CDB-Man has joined #gnucash
01:41:23 *** CDB-Man_ has quit IRC
02:32:36 *** MechtiIde has joined #gnucash
03:02:04 *** MechtiIde has quit IRC
03:05:58 *** MechtiIde has joined #gnucash
03:29:31 *** gour has joined #gnucash
03:48:02 *** gour has quit IRC
03:52:05 *** gour has joined #gnucash
04:28:25 *** MechtiIde has quit IRC
04:42:03 *** rubdos has joined #gnucash
05:29:31 *** fabior has joined #gnucash
05:48:52 *** thomasr has joined #gnucash
05:53:13 *** thomasr has quit IRC
05:53:53 *** fabior has quit IRC
06:47:56 *** MechtiIde has joined #gnucash
06:49:41 *** AndBill has joined #gnucash
06:57:35 *** thomasr has joined #gnucash
07:12:00 <mikee> thomasr: The customer importer only checks for duplicate IDs. No checking for duplicate names or email. Improvements are welcome.
07:13:40 <mikee> See: src/plugins/customer_import/dialog-customer-import.c near lines 347
07:22:17 *** Jimraehl1 has left #gnucash
07:26:19 *** Jimraehl1 has joined #gnucash
07:36:53 *** fabior has joined #gnucash
07:50:01 *** ErKa has joined #gnucash
08:28:16 *** Coderjoe has quit IRC
08:28:25 *** fabior has quit IRC
08:31:15 *** Coderjoe has joined #gnucash
08:47:33 *** ErKa has quit IRC
08:48:05 *** ErKa has joined #gnucash
09:25:12 *** ErKa has quit IRC
09:47:31 *** fabior has joined #gnucash
10:04:15 *** uXus has quit IRC
10:07:27 *** uXus has joined #gnucash
11:22:03 <Bill> Browsing Gnome Bugzilla, it seems Gnucash is desperate for some developers. Am I looking in the wrong place or misinterpreting?
12:17:17 <warlord> Bill: looking in the wrong place for ..???
12:18:10 <warlord> It is certainly true that the Developer to User ratio is rather low. But we do have active development. If you want to become a developer, we welcome you. If you're asking "is GnuCash dead" the answer is an unequivocal "no"
12:55:50 <Bill> warlord, it doesn't appear to be dead, and as a long time user, I sure hope not. But...
12:56:36 <Bill> I tried to use latest from repos in Ubuntu and a very very basic operation (duplicate simple transaction) crashes it. Very worrying...
12:57:00 <Bill> That sent me to the bugzilla were I see dozens of crashes noted that are years old.
12:57:10 <Bill> So... trying to get oriented.
13:02:38 <Bill> Now I've got the latest source from github and building is not going smoothly. Dependency problems.
13:03:40 <warlord> Bill: are you trying to build master or maint?
13:05:49 <Bill> git branch shows master
13:06:29 <warlord> that explains it -- master is the (unstable) dev branch, and yes, it has many more dependencies. You should build maint
13:06:36 <warlord> (which is the 2.6.x branch)
13:06:44 <Bill> It is unhappy with absence of libtoolize, intltoolize, automake, autoconf...
13:06:53 <Bill> still mucking with it
13:07:04 <Bill> OK, I'll try maint
13:07:39 <warlord> Bill: Umm, that soundslike you dont have a build environment!
13:07:45 <warlord> try: apt-get build-dep gnucash
13:08:28 <Bill> Yes, it does, but obtaining the missing components was not going well. Thanks I'll do that
13:09:32 <MechtiIde> I listed the build dependencies last week in wiki.gnucash.org/wiki/De/Übersetzung
13:09:41 <MechtiIde> for building in Debian
13:10:15 <MechtiIde> and then it works smoothly
13:11:05 <MechtiIde> Bill, ^
13:12:17 <Bill> Thank you, MechtiIde - it seems the build-dep worked very well. Got through autogen.sh now.
13:16:14 *** nomeata has joined #gnucash
13:21:00 <jralls> warlord, Bill: The only extra hard dependency in master is boost. There's also a soft dependency on google test and google mock for some of the tests, but configure figures out if they're installed and skips those tests if they're not.
13:23:45 *** AndBill has quit IRC
13:24:29 *** gour1 has joined #gnucash
13:25:13 *** gour has quit IRC
13:31:10 <Bill> Thanks, jralls, I stepped away. I'll keep that in mind. Just puzzling out which configure options make sense for me.
13:50:17 <warlord> Bill: other than --prefix=/opt/gnucash, you probably dont need any
13:53:26 <Bill> warlord, that's what I settled on. I chown'ed to me and ran till the boost dependency was noted
13:53:34 <Bill> looking into that now
13:56:23 *** fabior has quit IRC
13:57:48 <jralls> Ah, you didn't switch to the maint branch, then. Just apt-get install libboost-dev should do it.
13:59:10 <jralls> And apt-get install google-mock will install both google test and google mock.
14:01:30 <Bill> Doh! OK, too easy. I had just unpacked boost source, but that was nicer.
14:02:43 <Bill> Dang! swig "too old" need at least 2.0.10
14:03:28 <Bill> apt-get swig fixes that.
14:03:50 <Bill> configure is happy now.
14:07:33 <Bill> But make fails in ld... cannot find -lboost_date_time
14:07:41 <jralls> Bill: BTW, if your goal is to help with the bugs, we do that on the maint branch. You might find http://wiki.gnucash.org/wiki/Development and the pages that link from it helpful.
14:09:16 <Bill> I'd like to but I can't really commit to that now. Just being a parasite for the moment.
14:10:06 <Bill> If I can get this running, I'll trying switching ot maint, and see if I can get my head into it.
14:10:19 <Bill> Have a fiscal year to get to the tax guy ATM.
14:13:24 <jralls> Bill: Don't do that with master! Derek was serious when he said that it's unstable.
14:15:32 <Bill> K, jralls - but I just want to make something run from source first.
14:16:07 <Bill> needed apt-get install libboost-date-time-dev too
14:16:08 <jralls> OK. No guarantees that master will run...
14:16:35 <Bill> Understood. Need to get through this orientation.
14:16:47 <jralls> That's odd. libboost-dev should have installed everything.
14:17:10 <Bill> Nope - not from the Ubunutu repo.
14:17:28 <Bill> There are a lot of libboost*-dev thingies in the repo.
14:17:53 <Bill> make is chugging through now. So far clean...
14:19:00 <Bill> Ouch. -Wno-deprecated-register is unrecognized command line option.
14:21:53 <jralls> It should only say that when there's another real error. Look a couple of lines up.
14:22:13 <jralls> Or switch to clang. ;-)
14:23:08 *** MechtiIde has quit IRC
14:24:30 <Bill> Oh, duh, yeah, massive bitching, beyond scrolling...
14:24:58 <Bill> Switch to clang? Now I'm feeling really dumb...
14:28:28 <warlord> Bill: switch to maint branch.. seriously
14:29:41 <Bill> warlord, OK, but what about clang? Recommendation. I haven't had my head in C for 15 years, and now I see another part of the world was built...
14:29:51 <Bill> ?
14:30:13 <jralls> Bill: Note the ;-). Clang is a different compiler that implements -Wno-deprecated-register so you won't get that error. It won't do anything about the rest of the errors, though.
14:31:25 *** MechtiIde has joined #gnucash
14:32:09 <Bill> OK, I get the joke now. :-)
14:32:42 <Bill> But clang / LLVM looks interesting. Wonder how it escaped me.
14:32:53 <jralls> Though it also has the advantage of having much better error messages with C++ templates. If you've got hundreds of lines of spew that's probably why.
14:33:03 <jralls> A template error, that is.
14:34:52 <Bill> Yes, it was template related/
14:38:03 <Bill> maint branch: Tripped at no rule to make target gnc-date.cpp
14:38:18 <Bill> I'm going to wipe it out and start again from the git glone.
14:38:21 <Bill> clone
14:40:14 <warlord> yeah, good idea.
14:40:19 *** thomasr has quit IRC
14:41:24 <jralls> Nah, just do git clean -fdx.
14:42:34 <jralls> But also don't build in the source directory. Build in a different directory that you can rm -rf * in.
14:42:41 <Bill> jralls, never knew about that. I tried make clean, but NG. I've got a fresh copy of maint from github now. Let the machine sweat.
14:43:09 <Bill> Do you mean --prefix of configure?
14:43:20 <jralls> ?
14:43:56 <Bill> Don't build in source?
14:44:05 <Bill> I'm not thinking that deeply. ...
14:44:06 <warlord> What I do is use lndir to create a symlink (build) tree and then build in the symlink'ed tree.
14:44:14 <Bill> Just following steps from wiki
14:44:56 <Bill> OK. Well, make running now anyway.
14:45:09 <Bill> I'll see what happens
14:48:29 <jralls> The easiest way is mkdir build && cd build, then ../autoconf.sh && ../configure --prefix=foo && make V=0. Then you can rm -rf * in build without blowing up the rest.
14:49:09 <jralls> make distclean should also get rid of .deps, which is what is the source of the "can't find foo.cpp" when you switch branches.
14:50:06 <Bill> jralls, thanks - noting that so I can consider for future. Follow concept but not precise particulars.
14:50:53 <Bill> Anyway - make looks good so far. Fingers crossed.
14:54:46 <Bill> make succeeded.
14:57:24 <Bill> And the sucker runs without crashing like the one from the Ubu repo crashes.
14:57:57 <Bill> Thank you, warlord, jralls
14:58:19 <jralls> You're welcome. Have fun with those taxes!
14:58:27 <Bill> Ecch
14:58:49 *** fell__ has joined #gnucash
14:59:23 <warlord> Glad it worked.
15:00:38 *** fell_ has quit IRC
15:10:16 *** MechtiIde has quit IRC
15:17:17 *** MechtiIde has joined #gnucash
15:20:52 <Bill> I would like to add some functionality to it, but I think it might take me too long to get into the code deep enough to do anything well.
15:27:36 *** MechtiIde has quit IRC
15:30:08 *** karelk has joined #gnucash
15:33:15 <jralls> Bill: GnuCash is indeed complex. We're also quite insistent upon keeping it focussed on accounting. If you need non-accounting features like inventory control, consider a different package.
15:37:49 <Bill> I think that is appropriate.
15:38:35 <Bill> In support of that separation, I would like to be able to dump transactions into it from other processes.
15:38:59 <Bill> Perhaps if it ran in a server mode listening to a socket, with a submission API.
15:40:14 <Bill> Although, all I want to do is spit stuff out of shell scripts when I get an invoice in the [e]mail.
15:41:01 <Bill> Or stuff like that.
15:47:31 *** thomasr has joined #gnucash
15:47:43 <jralls> In it's current state that will be very difficult. There's too much "business logic" in GUI files. One of our goals for the next couple of major releases is to clean that up.
15:54:32 <Bill> I suspected that. It's a beautiful GUI, but it's a PIA to haul it out and type into it for day-to-day stuff, if you're not a full time bookkeeper.
15:55:08 <Bill> Wanna just say, the phone bill was $3 this month from the command line and be done with it.
15:55:48 <Bill> But of course, this functionality would also support interoperation with other programs.
15:56:20 <Bill> I imagine adapting the schdeuled transaction templates for this purpose. Maybe parameterize them.
15:58:37 <jralls> I think that interchange with QIF or OFX would be more likely to succeed.
16:00:13 <Bill> That makes sense to me in that it uses an established standard.
16:00:39 <Bill> ...for data interchange...
16:01:18 <Bill> But I'm wondering how it provides a pretty guide to defining typical transactions for a given users, purpose.
16:01:27 <Bill> user's
16:04:25 *** nomeata has quit IRC
16:07:06 <jralls> It doesn't, but neither do scheduled transactions in their stored state. You'd need to write a wrapper that would take the pretty input and turn it into an interchange file.
16:15:38 <Bill> Yeah, that's about what I have in mind. I was only using scheduled transactions as a model because of their UI.
16:18:09 <Bill> But it would only a tool generating a transaction validation specification. If that makes sense.
16:18:19 <Bill> only be
16:18:30 <Bill> only be a tool for
16:23:37 <jralls> Making a valid OFX or QIF transaction would also validate the transaction.
16:23:40 <Bill> I suppose I could just write scripts to generate qfx batches and bring them into the program in batches now.
16:24:23 <Bill> Is that possible?
16:24:34 <Bill> I assume so, but never tried it.
16:24:40 *** rubdos has quit IRC
16:26:03 <jralls> Yes, that's basically what I've been suggesting.
16:26:10 <jralls> I've got to go now.
16:27:42 <Bill> OK, thanks again for your help.
16:40:01 *** thomasr has quit IRC
16:50:31 *** gour1 has quit IRC
17:33:50 *** mlncn_ has joined #gnucash
17:54:55 *** fell__ is now known as fell
17:55:57 *** gncbot sets mode: +o fell
18:14:44 *** mlncn_ has quit IRC
20:26:01 *** c_korn has joined #gnucash
20:27:09 <c_korn> hey there. the build of gnucash-docs 2.6.10 fails: make[4]: *** No rule to make target ``ls', needed by `gnucash-help.pdf'. Stop. http://zulu309.server4you.de/build/logs/gnucash-docs_2.6.10-1~getdeb1-i386-20151219-2216
20:37:48 <warlord> c_korn: considering all the crap on that page, I would suggest it's a build-server issue. We build the gnucash-docs nightly (well, whenever there is a change) and it's not failing here.
20:40:02 <warlord> c_korn: I just manually re-ran the nightly build in that directory and cannot reproduce your issue.
20:44:54 <c_korn> warlord: I can reproduce it here
20:45:58 <c_korn> maybe this file in the Makefile causes the problem: figfiles = `ls $(figdir)/*.png`
20:48:21 <warlord> Hmm... I do not see that in my Makefile. I see: figfiles = $(figdir)/*.png
20:50:59 <warlord> Hmm... Looks like a difference between maint and master. I DO see that in maint (looking at github)
20:51:41 <warlord> jralls: should the server be building maint or master of gnucash-docs (and other docs)??
20:53:38 <c_korn> this commit seems to introduce the issue https://github.com/Gnucash/gnucash-docs/commit/254029ab75db44a3634a67e274a20ed86d72aa7a
20:55:48 <warlord> Hmm.. are you saying that you don't have an 'ls' in your PATH?
20:56:30 <c_korn> there is /bin/ls in the PATH
20:56:44 <warlord> Arguably he should have used $shell(ls ...)
20:57:04 <warlord> ala: gnomehelpfigures_DATA = $(shell ls ${srcdir}/${figdir}/*.png)
21:13:05 <c_korn> just reverted that commit and it works. thanks.
21:13:07 *** c_korn has quit IRC
22:38:49 *** jethrogb has quit IRC
23:59:30 *** volan has joined #gnucash