2014-04-29 GnuCash IRC logs

00:14:02 *** fell has joined #gnucash
00:14:03 *** gncbot sets mode: +o fell
00:56:26 *** MechtiIde has joined #gnucash
01:08:27 *** ErKa has quit IRC
01:40:33 *** MechtiIde has quit IRC
01:41:11 *** fell has quit IRC
02:32:58 *** O01eg has quit IRC
02:42:23 *** sean76 has joined #gnucash
02:43:55 <sean76> Using GnuCash 2.6.3 latest download windows version. Getting Unable to retrieve quotes for these items error.
02:44:10 <sean76> already installed perl module successfuly
02:44:17 <sean76> any help?
02:46:10 <sean76> I am using standard stock such as Amazon (AMZN) in nasdaq
02:47:14 <sean76> using windows 7 pro.. if thats needed
02:56:22 *** gjanssens has joined #gnucash
02:56:22 *** gncbot sets mode: +o gjanssens
03:32:22 *** benoitg has joined #gnucash
04:14:59 *** floh1111 has joined #gnucash
04:16:28 *** floh1111 has quit IRC
04:47:37 *** gjanssens has quit IRC
04:51:03 *** gjanssens has joined #gnucash
04:51:03 *** gncbot sets mode: +o gjanssens
04:59:17 *** StuM has quit IRC
05:09:19 *** gjanssens has quit IRC
05:32:58 *** gjanssens has joined #gnucash
05:32:58 *** gncbot sets mode: +o gjanssens
05:48:31 *** Ardonik has quit IRC
05:56:08 *** Ardonik has joined #gnucash
09:09:54 *** gncbot has joined #gnucash
09:10:01 <gjanssens> But I'm afraid that's not may area
09:10:14 <gjanssens> yay gncbot returned :)
09:10:52 <gjanssens> Let's all call out for warlord now :D
09:10:59 <lmat> :)
09:11:08 <lmat> gjanssens: I'm reading http://gnucash.org/download.phtml
09:11:17 <lmat> gjanssens: What distribution do you use ?
09:11:26 <gjanssens> Fedora
09:12:10 <gjanssens> lmat: you have no way of getting your debian install online ? That would simplify things a lot...
09:12:16 *** gjanssens sets mode: +o gncbot
09:13:14 <lmat> gjanssens: Nothing practical. Besides, I would much rather be able to do it offline :)
09:13:17 <lmat> gjanssens: At home, I don't have internet access.
09:13:43 <lmat> gjanssens: apt-get --print-uris "Instead of fetching the files to install, their URIs are printed."
09:13:48 <lmat> gjanssens: Sounds perfect :)
09:14:05 <gjanssens> lmat: indeed
09:14:25 <lmat> gjanssens: apt-get install --print-uris gnucash | wc returns 64
09:15:31 <lmat> apt-get build-dep --print-uris gnucash | wc -l returns 185
09:16:25 <gjanssens> lmat: if you don't intend to build gnucash yourself it's probably sufficient to install the 64 from your first command
09:21:51 <lmat> gjanssens: I am a developer, and interested in contributing, or at least fiddling :)
09:22:07 <gjanssens> lmat: you're mostly welcome :)
09:22:13 <lmat> gjanssens: Although I am averse enough to Object-Oriented C (especially the way gnucash does it), that I would probably just avoid it altogether.
09:22:39 <gjanssens> You could help jralls convert it to proper c++ is you like
09:22:59 <gjanssens> And I meant "most", not "mostly"
09:23:03 <lmat> gjanssens: That's what I like! I suggested that to warlord, and he pointed out the many hundreds of man hours that have been spent on gnucash.
09:23:21 <lmat> gjanssens: heh, I didn't catch that ^_^
09:23:32 <lmat> gjanssens: Thanks for pointing out an effort in that direction.
09:23:41 <lmat> jralls: yo! Where is your c++ gnucash code?
09:23:44 <gjanssens> I'm not a native English speaker so sometimes I make funny errors :)
09:24:05 <lmat> gjanssens: no problem, I'm much more graceful on IRC about that sort of thing ^_^
09:25:07 <gjanssens> lmat: the c++ effort has only just started
09:25:24 <gjanssens> You could look at the git history of the last couple of days
09:25:25 <lmat> gjanssens: It's about time! And that's just the place where I would like to get started: right at the beginning!
09:25:30 <lmat> gjanssens: yeah, where's that repo ?
09:25:56 <lmat> (thank God for free software!)
09:26:02 <gjanssens> lmat: https://github.com/Gnucash/gnucash
09:26:44 <gjanssens> And let me point you at our git usage guidelines as well: http://wiki.gnucash.org/wiki/Git
09:26:54 <lmat> gjanssens: thanks
09:27:12 <gjanssens> Btw it's probably a bit early to call for jralls (early as in early morning)
09:27:25 <lmat> gjanssens: Oh, so I'm looking for gnucash/gnc ?
09:27:38 <gjanssens> He usually pops up later even though his nick is still online
09:27:49 <gjanssens> gnucash/gnc ?
09:28:04 <lmat> gjanssens: He must be lazy, no "away" is set :-P
09:28:23 <lmat> gjanssens: Yeah..., I'm seeing "main.cpp -- The program entry point for cutecash"
09:29:03 <gjanssens> Actually no that's not the focus of the c++ effort
09:29:11 <lmat> gjanssens: Oh, that's the QT version.
09:29:16 <lmat> gjanssens: Where in the repo is the c++ effort ? ^_^
09:29:17 <gjanssens> cutecash was an initial experiment
09:29:29 <gjanssens> jralls started in libqof/qof
09:29:47 <gjanssens> The first step was only to make the existing c code compile as c++
09:29:50 <lmat> gjanssens: seems like I used to know what qof is...
09:29:58 <lmat> gjanssens: Oh
09:30:33 <gjanssens> The next step was to introduce boost::guid to replace the gnucash guid code
09:30:46 <lmat> gjanssens: ahh
09:31:02 <gjanssens> jralls probably works on a private branch somewhere
09:31:07 <lmat> gjanssens: yeah.
09:31:15 <lmat> gjanssens: I'll need to grab some of his commits ^_^
09:31:16 <gjanssens> So far he's the only one to actively work on the c++ part
09:31:38 <lmat> gjanssens: Well, since the c++ port will perform better than the original, I think I'll get on board :-P
09:32:26 <gjanssens> Great! I think we should arrange for more public tracking of this work then.
09:32:38 <gjanssens> Talk to jralls when he shows up on irc
09:33:09 <lmat> gjanssens: Do you prefer c or c++ (or do you program?)
09:33:44 <gjanssens> That is hard to explain in 2 words :)
09:33:50 <gjanssens> I would prefer c++
09:34:05 <gjanssens> However my last bit of c++ I wrote was more than 15 years ago
09:34:34 <gjanssens> I started contributing to gnucash because I needed some extra features and wanted to do some programming again
09:35:05 <gjanssens> It took me several years to get used to the OO semantics in C and gobject
09:35:32 <gjanssens> I'd be happy to switch to real OO in c++ but I'm well awar that my c++ knowledge is horribly outdated
09:36:01 <gjanssens> So I'll just passively follow the conversion for now and learn as much as I can from the work of others :)
09:36:56 <lmat> gjanssens: hah! okay
09:37:00 <gjanssens> Meanwhile I'll make myself useful in other ways
09:37:19 <gjanssens> Like revamping the windows build scripts, fixing bugs, polishing the ui,...
09:37:26 <lmat> gjanssens: I see. I bit the bullet and learned c++ a couple months ago.
09:38:24 <lmat> gjanssens: I learned the c++0x standard, so I'm fairly up-to-date (using universal initialization, and other new-fangled features ^_^)
09:38:31 <gjanssens> lmat: how did you learn it ?
09:38:41 <lmat> gjanssens: Oh yeah, that's the only standard I know... I wonder what jralls is planning to use!
09:38:47 <lmat> gjanssens: I went to ##c++ ...
09:38:55 <lmat> gjanssens: And I wrote a few programs.
09:38:59 <gjanssens> c++0x ^_^
09:39:09 <lmat> gjanssens: He is ?
09:39:32 <gjanssens> Yep, that's what he said in a conversation on the ml a couple of days back
09:39:38 <lmat> git://git.code.sf.net/p/lawsa/gitrepo is mostly c++0x
09:39:48 <lmat> gjanssens: perfect. Oh, I'm on the mailing list, I should probably go back and read some of those messages :)
09:41:53 <gjanssens> lmat: actually it was a conversation on irc, not on the mailing list
09:42:01 <gjanssens> Here's the link to the logs: http://lists.gnucash.org/logs/2014/04/2014-04-26.html#T15:53:13
09:42:05 <lmat> gjanssens: :)
09:42:12 <lmat> gjanssens: Did you say the irc logs are on warlord's machine ?
09:43:15 *** wol has quit IRC
09:43:26 *** aqua___ has quit IRC
09:44:18 <gjanssens> Yes, which came back online together with gncbot
09:44:33 <gjanssens> So it likely was a network outage only, not a server meltdown :)
09:51:55 <lmat> gjanssens: Well, I'll probably forget, so if you remember, please introduce me to jralls ?
09:54:00 <gjanssens> lmat: I will if I'm around
09:54:20 <gjanssens> I won't be for long anymore today
09:54:29 <lmat> gjanssens: I see.
09:58:45 *** himaxx has joined #gnucash
09:59:23 *** warlord has joined #gnucash
09:59:34 <warlord> @op
09:59:36 <gncbot> warlord: Error: You don't have the #gnucash,op capability. If you think that you should have this capability, be sure that you are identified before trying again. The 'whoami' command can tell you if you're identified.
09:59:48 <warlord> @op
09:59:48 *** gncbot sets mode: +o warlord
10:00:41 <warlord> Hey all. Had a 3 hour power outage here. UPS only lasted a bit over 2 hours.. but of course Comcrap buggered out at the first sign of a loss of power...
10:02:08 <gjanssens> warlord: welcome back
10:02:31 <gjanssens> I have entertained the crowd in your absence :)
10:02:41 <gjanssens> Well, I tried to answer what I could...
10:03:16 *** himaxx has quit IRC
10:06:34 *** jimvideo has joined #gnucash
10:07:03 <lmat> gjanssens has been an excellent host.
10:07:11 <warlord> Thanks.. :)
10:08:03 <warlord> sean76: what quote source? And what version of F::Q do you have installed? It's possible (based on the perl package you installed) that you don't have the most up to date version to handle recent Yahoo website changes.
10:10:01 <warlord> Anyways, code and win32 should all be back online.. I had one VM service not come back; I've just fixed that..
10:16:15 <sean76> guys, I am new to gnucash and i am not able to get the online quotes. I have installed the perl Finance:: module
10:16:41 <sean76> i am using gnucash latest version on windows 7 pro
10:16:57 <warlord> sean76: yes, we saw your question -- you don't have to ask it again.
10:17:09 <warlord> What version of F::Q do you have? And which version of PERL did you install?
10:18:00 <sean76> it should be latest.. i installed only yesterday... i dont know the version..
10:18:26 <warlord> What about perl?
10:18:32 <warlord> which perl did you install?
10:19:10 <sean76> how can i find it?
10:19:37 <warlord> No clue.. I don't use Winblows
10:20:46 <warlord> Did you download strawberry perl or active perl?
10:20:57 <sean76> perl 5 version 16
10:22:37 <sean76> active perl
10:22:55 <sean76> that was in the documentation
10:23:32 <sean76> and ran Install Online Price Retireval
10:23:42 <sean76> from start menu
10:23:54 <warlord> gjanssens: isn't active perl the version that does not have current F::Q? I've lost track.
10:24:11 <gjanssens> yes
10:24:38 <gjanssens> Last time I heard only Strawberry perl has a working F::Q
10:24:54 *** ErKa has joined #gnucash
10:24:58 <gjanssens> But maybe this has changed by now it's been a couple of weeks
10:25:11 <warlord> sounds like not based on sean76's statements.
10:25:33 <sean76> hmm.. if thats the case the documentation must be updated
10:26:16 <warlord> I thought current (2.6.x) installer script would install perl automagically if it wasn't already there. gjanssens ?
10:27:43 <gjanssens> warlord: that should be the case indeed
10:28:04 <sean76> no it did not autmatically install.. i had to go to http://www.activestate.com/activeperl
10:28:04 <gjanssens> It probably the documentation needs to be updated indeed sean76
10:28:46 *** aqua___ has joined #gnucash
10:28:57 <sean76> i am going to install strawberry perl lets see
10:30:08 <gjanssens> sean76: did you get an error message when running install-fq-mods.cmd ?
10:30:34 <sean76> no.. got successfully installed message
10:30:39 <gjanssens> Perhaps we need to update the version of strawberry to download if the old version is no longer online
10:31:00 <gjanssens> And did you install perl yourself before running install-fq-mods.cmd ?
10:31:34 <sean76> yes..
10:32:24 <gjanssens> Then install-fq-mods.cmd won't attempt to install strawberry
10:32:41 <gjanssens> It will only do so if it doesn't find any perl
10:32:57 <sean76> oh..
10:33:02 <gjanssens> You may want to uninstall activeperl and the run install-fq-mods again
10:33:12 <sean76> ok
10:37:28 *** gjanssens is now known as gjanssens_
10:46:18 <sean76> it worked.. thanks
10:46:33 <sean76> so strawberry perl must be installed..
10:46:34 <warlord> Yay
10:46:53 <sean76> please update documentation when you get a chance..
10:47:51 *** sean76 has quit IRC
10:48:25 *** O01eg has joined #gnucash
11:12:25 <warlord> gjanssens_: two recent changesets didn't have diffs when sent to -changes. Any idea why the script didn't include diffs?
11:54:34 *** aqua___ has quit IRC
11:57:54 *** benoitg has quit IRC
12:07:34 <jralls> warlord: Looks like it's been that way since I changed the script April 12th. I use gnucash-patches instead of gnucash-changes, so I hadn't noticed.
12:12:13 <warlord> jralls: Hmm
12:13:23 <jralls> I'm looking at it now, trying to figure out why my changes would have affected the output when the change isn't a new branch.
12:13:43 <warlord> jralls: thanks!
12:34:08 *** StuM has quit IRC
12:35:54 <gjanssens_> jralls: it has something to do with merges vs direct branch commits
12:36:43 <jralls> I don't think so. Your change was a merge, but Christian's wasn't.
12:36:48 <gjanssens_> I seem to remember the script won't print diffs for merges assuming the diffs had already been reported earlier
12:37:22 *** MechtiIde has joined #gnucash
12:38:13 <gjanssens_> Hmm, indeed.
12:38:37 <gjanssens_> Oh well, merges may be part of this issue
12:39:11 <gjanssens_> jralls: before I dash again, lmat is very interested in helping in the conversion to c++
12:39:40 <gjanssens_> perhaps the two of you can check out how he can help
12:39:58 * gjanssens_ is off again...
12:40:10 <jralls> OK. I got an email yesterday from a guy named Sumit offering the same. I asked him to reintroduce himself on the dev list.
12:40:12 *** jmd has joined #gnucash
12:41:26 *** gregg4567 has joined #gnucash
12:47:50 *** jmd has quit IRC
12:48:15 *** jmd has joined #gnucash
13:11:33 <jralls> warlord: This is weird. The fix locally seems to be to remove the --stdin argument from `git rev-list` when generating the list of commits to show. I'll push it and see what happens.
13:12:21 <warlord> ok.
13:13:45 <jralls> lmat: Geert says you want to help with C++. Do you have anything in particular in mind?
13:22:06 *** ErKa has quit IRC
13:36:26 *** Bodhi-Baum has joined #gnucash
13:37:20 *** TradeBorG11111 has joined #gnucash
13:43:16 <lmat> jralls: No.
13:45:19 <jralls> Umm, OK. Do you want me to assign you something? Are you up to speed on C++11?
13:46:11 <lmat> jralls: Yeah, gjanssens_ said that you have been working on qof. I think so ^_^
13:47:12 <jralls> Qof is the root of everything else, so that's where I started. But all I've done so far is to make it compile as C++ instead of C.
13:47:31 <jralls> And that's about where I am with C++11, too. ;-)
13:49:32 <lmat> jralls: Okay, do you have a grand plan, or are you going one code file at a time ?
13:52:30 <jralls> Both. My plan is to replace the implementation of GncGUID with boost::guid, of gnc-numeric with boost::rational<boost::multiprecision>, and of gnc-date with boost::calendar::gregorian and boost::datetime. (boost module names from memory, so I might have misspelled some of them.)
13:53:16 <jralls> Mike A wants me to get gnucash-bin.c converted to gnucash-bin.cpp along the way. You probably saw that on the dev list yesterday.
13:54:19 <jralls> But one at a time, tested on all three platforms including make check, and merged to master.
13:55:34 <lmat> jralls: merged to master ? So gnucash is "officially" taking this on ?
13:55:41 <jralls> After that, start rewriting some of the more independent classes in QoF as C++11; maybe qofsession and qofbackend.
13:55:51 <jralls> Oh, absolutely.
13:56:02 <lmat> jralls: hmm, good to know.
13:56:42 <jralls> The long term goal is to rewrite the engine without glib, and especially without gobject.
13:56:56 <lmat> jralls: That was my goal the first time I looked at the code ^_^
13:57:45 <lmat> jralls: Well... I'm trying to think of where I should set foot first.. I guess I should try to compile using a c++ compiler first ? Are there any patches that have been done that I'll need ? Should I have a reference to your git repo ?
13:58:24 <jralls> If you've built master since yesterday you already have. ;-)
13:59:10 <lmat> jralls: I see! (no I haven't)
13:59:29 <jralls> I merged it Friday afternoon and Mike A, Geert, and I spent much of Saturday fixing stuff because I hadn't cross-tested properly.
13:59:42 <lmat> jralls: Gotcha
14:00:14 <jralls> Speaking of which, what's your setup? Do you have VMs for cross-testing?
14:00:44 <lmat> jralls: I'm on ubuntu, g++ 4.8.1. I have virtual machines, but not for cross-testing ^_^
14:01:50 <jralls> No Windows, eh? Are you willing to set one up?
14:02:25 <lmat> jralls: Unfortunately, Microsoft Windows isn't free software, so willing isn't the end of the questioning ^_^
14:02:37 <lmat> jralls: I happen to have Windows 7 installed as a virtual machine (where I use Outlook for work)p
14:04:09 <jralls> Well, "willing to pay" is a subset of "willing". Can you use that VM, perhaps as a different user for isolation, or did your employer pay for it and impose restrictions?
14:05:12 <lmat> jralls: I don't think there are restrictions on its use. Different user sounds like a good idea.
14:09:36 <jralls> So I guess the first thing for you to do is to build master. I've got a mostly unrelated merge that I'm working on right now, a left-over from something I decided wasn't ready to put into 2.5.x 3 weeks before releasing 2.6. It moves all of the KVP manipulation inside of engine.
14:09:56 <lmat> jralls: Okay, I'm working on that now...gathering build dependencies.
14:10:49 <jralls> Have you done any profiling? If so, what tools do you like to use?
14:11:20 <lmat> jralls: I've only used http://en.cppreference.com/w/cpp/chrono/high_resolution_clock for profiling
14:12:30 <lmat> jralls: Sorry, I haven't done a lot of the "normal" stuff like use a debugger, etc.
14:12:47 <jralls> So you wrap a function call in that and see how much time it takes?
14:13:04 <lmat> jralls: something like that.
14:13:23 <jralls> You code so well the first time that you never need a debugger? Either awesome or nuts...
14:13:35 <lmat> jralls: I use logging to get through problems.
14:17:20 * jralls nods. Seems slow.
14:20:26 <jralls> What about valgrind? Have you ever used that?
14:20:33 <lmat> jralls: I did, just recently!
14:20:44 *** ErKa has joined #gnucash
14:20:48 <lmat> jralls: Someone suggested it because I had an out-of-bounds, and I didn't want to add the necessary logging statements :)
14:21:44 <jralls> That's one of the things that logging statements aren't much help with. Double-frees are another.
14:23:00 <jralls> Then there's problems like https://bugzilla.gnome.org/show_bug.cgi?id=724590. I can't imagine how I'd find that with logging.
14:24:44 <lmat> jralls: I had assumed that a vector contained two elements (it usually does in this context), but it only had one :(
14:26:12 <lmat> jralls: Ahh, I see the slab_stack corruption problem. Good catch :D
14:28:05 <jralls> I'd have trapped on the exception in the debugger and then looked for why the vector was undersized. I'd think that would be easier than looking at allocations, which can be automatic for std::containers.
14:30:39 <lmat> jralls: All I wanted was a call stack. I got that from valgrind. Once I had that, I was able to read the code and see the error.
14:31:44 <lmat> I'm running 'make' in gnucash, and my interpreter is barfing on src/engine/iso-currencies-to-c at the line 'exec @GUILE@ ...' I don't think "@GUILE@" is some kind of sh variable syntax... is something supposed to expand that first ?
14:33:54 <jralls> That's autotools substitution syntax. Just a minute, I'll look where it's supposed to be subbed.
14:40:00 <jralls> Configure should have subbed that.
14:40:31 <jralls> Do you have the appropriate guile-dev package installed?
14:40:53 <lmat> jralls: I think so. configure complained about it, then I installed, then configure didn't complain anymore. I'll take another look and make sure I got it right.
14:41:54 <lmat> jralls: looks like probably not.
14:43:16 <lmat> I have guile-1.8{,-dev,-libs} ...
14:45:28 <lmat> (the exact error is "GUILE_PROGS" not found)
14:47:25 <warlord> gjanssens_: I just got your email ;)
14:47:31 <lmat> GUILE_PROGS only shows up once in the configure file. I had assumed that configure defines it, but I guess not!
14:47:35 <warlord> lmat: sounds like GUILE isn't there?
14:47:40 <lmat> I have guile-1.8{,-dev,-libs} ...
14:48:10 <warlord> lmat: @GUILE@ should be filled in by the results from AM_FIND_PROG(GUILE, guile) -- or the equivalent.
14:48:15 <jralls> What exact version of guile-1.8? GC requires gnuile-1.8.5 or later.
14:48:16 <warlord> It sounds like configure isn't finding guile?
14:48:38 <lmat> warlord: The only place GUILE_PROGS is found is after configure decides that I do have guile. My version is...
14:48:44 <lmat> 1.8.8
14:49:00 <jralls> Should be OK.
14:49:47 <warlord> what's the full actual error you get?
14:50:25 <lmat> When running ./configure --prefix /home/lmat/sources/builds/gnucash , I get
14:50:34 <lmat> "./configure: line 20533: GUILE_PROGS: command not found"
14:50:48 <lmat> This is after "checking for guile-1.8 >= 1.8.5... yes" and some others.
14:51:00 <lmat> oh, there's another warning about my swig being too old..
14:51:53 <warlord> okay, so what, exactly, is line 20533?
14:52:02 <jmd> [the number of dependencies of gnucash is obscene]
14:52:35 <lmat> jmd: Rewrite the reporting facilities in c++ ^_^
14:52:44 <warlord> Ewwwww
14:52:45 <warlord> No...
14:53:18 <lmat> warlord: 20532 says if test $succeeded = yes; then 20533 says GUILE_PROGS 20535 says else
14:53:43 <lmat> In short.... if test $succeeded = yes; then GUILE_PROGS; else; succeeded=no ...
14:53:56 <warlord> lmat: can you pastebin a bit more context around that?
14:53:56 <lmat> In english... if the guy has guile, run GUILE_PROGS (which should be defined by the system?)
14:54:02 <lmat> warlord: Yeah sure :)
14:54:18 <warlord> lmat: I understand shell -- no need to "translate" to me. thanks
14:54:42 <jralls> It's an m4 macro that should be defined by guile.m4, which in turn should have been installed by your guile-1.8-dev package.
14:55:23 <warlord> AHH!!!! There are some systems that DONT install the .m4 file in the guile-18 dev package but only in the guile2 dev package!
14:55:40 <lmat> sheesh ^_^
14:55:52 <lmat> That pastie, for the record, is http://pastie.org/9124993
14:56:22 <lmat> Okay, so I'll install guile-2 ?
14:56:40 <jralls> You can, if you like.
14:56:52 <warlord> we do support both 1.8 and 2.0
14:57:20 <jralls> ISTR that Geert told someone the other day to uninstall guile-1.8 first.
14:57:31 <lmat> I did :)
14:58:12 <lmat> "error: guile 2 is found on your system, but appears to export different autoconf macros than what we expected. Please report this as a bug in GnuCash, so we can fix this for your platform." ^_^
15:01:25 <jralls> So neither package on your distro has GUILE_PROGS defined. Great.
15:01:34 <lmat> jralls: I don't know about that.
15:01:42 <lmat> jralls: the guile 2.0 error I don't understand. it seems very vague to me.
15:02:10 <lmat> jralls: I think with guile 2.0, it didn't got so far as GUILE_PROGS. But could you send me a guile.m4 from 1.8 ? ^_^
15:02:36 <jralls> That's because you're not looking at gnucash/configure.ac line 463, which is where that error message is emitted when it can't find GUILE_PROGS.
15:02:49 <lmat> jralls: yup, that's why.
15:03:01 <jralls> Happen to be looking at a copy at http://stuff.mit.edu/afs/sipb/project/guile16/share/aclocal/guile.m4
15:03:07 <lmat> hah
15:03:41 <warlord> FWIW, that line in my configure says:
15:03:42 <warlord> if test $succeeded = yes; then
15:03:42 <warlord> # Extract the first word of "guile", so it can be a program name with args.
15:03:55 <lmat> so I'll stick that in /usr/share/aclocal...
15:04:25 <warlord> lmat: honestly we should really fix the issue so it actually works
15:04:42 <jralls> Oh, time for lunch with the spousal unit. Back in a while.
15:05:20 <lmat> warlord: guile was already in /usr/share/aclocal.
15:05:40 <warlord> lmat: okay
15:06:42 <lmat> warlord: So we need to make sure configure sees it ?
15:06:48 <lmat> warlord: Does autogen.sh have anything to do with this?
15:07:10 <lmat> warlord: I ran autogen, then ran configure. Configure failed, so I installed something. Configure failed again, so I installed something else; rinse and repeat 12 times.
15:07:19 <lmat> warlord: Does autogen need to have guile handy to make everything work?
15:08:58 <lmat> I'm running autogen.sh again and it says "Running aclocal -I macros" I'm thinking that's important ?
15:09:32 <lmat> Heye, no guile warnings/errors this time. Moral: be sure to run autogen.sh before re-running configure ?
15:09:47 <lmat> But i do have a swig progrem... I need 2.0.10, not 2.0.4... should be easy enough to fix :)
15:13:55 <lmat> jmd: I think the build process seems very reasonable. But that may be because the last project I worked on was Weston ^_^
15:16:23 <lmat> warlord: Okay, I got it built (without the correct swig version), and running.
15:17:10 <lmat> warlord: I told it to start a new book. It did, and the title says "Unsaved Book - Accounts - GnuCash". Save is not available (presumeably because the only backend I installed was sqlite3, and it "auto-saves"). Now, how do I save my books ?
15:17:55 <lmat> It seems unintuitive to have to press "Save As..." in order to save my books for the first time, no ? Perhaps if the books are in the state "Unsaved Book", the save action should be available (even if using the sqlite3 backend, or any other backend that autosaves)?
15:18:06 <lmat> jralls: anyway, I guess I compiled successfully with gcc...
15:18:09 <lmat> jralls: I mean g++
15:23:47 *** wol has joined #gnucash
15:24:00 <warlord> lmat: How did you start a new book?
15:24:27 <lmat> warlord: new book, then cancel the wizard.
15:25:31 <warlord> You mean File -> New File...
15:25:40 <lmat> warlord: yeah :)
15:26:10 <warlord> If should give you the option to File -> Save
15:26:15 <warlord> (not just File -> Save As)
15:26:25 <warlord> there is no autosave before you actually save it somewhere.
15:27:23 <lmat> warlord: makes sense :)
15:27:39 <lmat> warlord: Maybe I didn't do it right... I'll be sure to confirm the behaviour before I bother you more about it :)
15:30:26 <warlord> Basically when File -> New File it closes the existing session and starts a new one but doesn't connect a backend.
15:30:40 <lmat> warlord: Ohhh... that makes sense.
15:31:42 <warlord> To connect you SHOULD be able to File -> Save... But it is possible that that wont be accessible and only File -> Save As is there. Arguably if you dont have a backend then File -> Save and File -> Save As could be the same action.
15:49:18 *** MechtiIde has quit IRC
16:05:12 *** jmd has quit IRC
16:12:21 <jralls> If you cancel the wizard then there's nothing to save until you at least create 1 account.
16:15:11 *** Bodhi-Baum has quit IRC
16:53:49 <lmat> jralls: I've been successfully building for a little while, and I'm taking a look at putting in the boost guid :)
16:54:03 <lmat> jralls: If you're working on that now, I should probably move to one of the next things you suggested.
16:57:12 <jralls> lmat: Haven't started yet, I'm still cleaning up that other merge. If you want to start, go ahead. Step 1 is to add a check for boost in configure.ac, but based on your issues with guile I suspect you're not too familiar with autotools, so I'll do that part after pulling your branch.
16:57:28 <jralls> Speaking of which, have you forked GC on GitHub?
16:58:10 <lmat> jralls: If GC is gnucash, I have a clone, and a "private" branch set up :)
16:58:44 <jralls> OK. Hmm. Hang on a sec.
16:59:51 <lmat> jralls: I already get the cutest error, "error: template with C linkage" :D (I checked and it's using g++ as expected, and I placed the include boost/uuid/uuid.hpp outside an extern "C" block) If you've seen this and know exactly what to do, I would appreciate your knowledge, otherwise, I'll continue to look
17:01:25 <lmat> jralls: Oh, duh... gnc-date.cpp is extern "C"{ when it includes qof.h which includes, which include, which includes... uuid.h
17:01:27 <lmat> jralls: oh well
17:02:31 <warlord> Heh
17:03:03 <lmat> I gues that's to be expected ^_^
17:03:29 <warlord> arguably the 'extern "C"' should happen as late as possible.
17:03:30 <jralls> Yes, all of the existing API needs to be c-linkage so that the rest of GC can use it. I want to use PImpl anyway, so make a new gncguid-impl.h and declare the new class there.
17:04:02 <lmat> warlord: heh, that's a lost cause, methinks!
17:04:11 <jralls> Then you can implement the member functions in the usual way in gncguid.cpp.
17:04:19 <lmat> jralls: sure.
17:05:37 *** wafflejock has quit IRC
17:08:17 <jralls> I just pushed my c++-build branch to my github repo (https://github.com/jralls/gnucash). If you drop your current fork and fork my repo and use that branch or branches rooted in it, it will make it easier for us to pass stuff back and forth before a section gets merged into master.
17:19:08 <lmat> jralls: Is https://github.com/jralls/gnucash different than https://github.com/Gnucash/gnucash ? The former is a "public mirror" of the latter, right ?
17:19:24 <warlord> lmat: the former is a fork of the latter.
17:19:36 <warlord> the latter is the official master "upstream" repo
17:19:49 *** TradeBorG11111 has quit IRC
17:19:59 <warlord> (technically the public mirror of code.gnucash.org, which technically is the master)
17:26:24 <lmat> I'm having a bit of trouble with git... I git remote add -f --tags jralls https://github.com/jralls/gnucash
17:26:37 <lmat> But it doesn't look like I got any of his tags...
17:26:48 <jralls> I don't have any tags.
17:27:06 <lmat> Right, I mean branches.
17:27:31 <lmat> so, git checkout c++-build; git checkout remotes/jralls/c++-build; , etc. all fail.
17:27:37 <jralls> Try `git branch -r`
17:27:38 <lmat> but git remote -v shows jralls
17:27:58 <lmat> jralls: It only shows 'origin' branches (which is set to the "official" gnucash url)
17:28:15 <warlord> lmat: did you 'fetch' from jralls to pull his tags/branches?
17:28:33 <warlord> "git remote" just sets up the remote; it doesn't pull any data
17:28:35 <lmat> warlord: I git fetch --all -t (and others
17:28:49 <lmat> warlord: the -f to git remote should fetch immediately :)
17:29:01 <lmat> warlord: and git fetch jralls, etc.
17:29:04 <warlord> the --tags wouldn't pull branches.
17:29:19 <lmat> warlord: hmmm maybe that's one of my problems. I'll read more
17:30:10 <lmat> warlord: removed --tags, like magic!
17:30:19 * warlord suggests just doing "git remote add <name> <uri>" and then "git fetch <name>"
17:30:39 <lmat> warlord: yup, yup. I need to remember the difference between a "ref" and a "tag" :(
17:31:03 <warlord> a tag is a type of ref ;)
17:35:18 <lmat> jralls: Okay, I moved my changes to c++-build.
17:35:54 <lmat> Gotta go now. It was nice to meet you, jralls . Talk to you tomorrow perhaps!
17:35:58 <warlord> yay rebase
17:36:25 <jralls> I'll be at Habitat tomorrow. See you Thursday.
17:36:39 <lmat> warlord: yeah, I didn't rebase this time :) I used cherry-pick to move my commit over :)
17:37:04 <warlord> jralls: fyi I'll be in SF/OAK M-Th
17:37:11 <lmat> okay, afk
17:37:16 *** lmat has quit IRC
17:37:50 <jralls> And I'm leaving for Richmond, VA on Sunday. Oh Well, missed another one.
17:38:30 <warlord> I'll be back
17:40:23 <warlord> ... certainly once or twice more thru july
17:41:29 <jralls> Oh, yeah, we're going to Belgium for July, too. Maybe we'll connect in June.
17:42:09 <warlord> *nods*
17:42:15 <warlord> I know I wont be out there in August
17:42:27 <warlord> *maybe* september
17:44:51 *** wafflejock has joined #gnucash
17:44:53 *** wol has quit IRC
17:45:32 <jralls> Well, unless you change jobs you're likely to be out here again sometime this fall. I don't have any more travel planned after July except for a wine-buying trip to Sonoma in August.
17:51:07 *** aqua___ has joined #gnucash
18:06:21 *** amjam has joined #gnucash
18:18:39 *** StuM has joined #gnucash
18:23:49 *** aqua___ has quit IRC
18:29:01 *** wafflejock has quit IRC
18:48:57 *** fell_ has joined #gnucash
18:48:57 *** gncbot sets mode: +o fell_
18:52:14 *** amjam has quit IRC
18:57:40 *** fell has quit IRC
19:39:20 *** wafflejock has joined #gnucash
19:51:45 *** gregg4567 has quit IRC
19:55:26 <warlord> jralls: exactly :)
21:14:43 *** hhn10 has quit IRC
21:26:51 *** gregg4567 has joined #gnucash
21:49:40 *** ErKa has quit IRC
23:25:44 *** StuM has quit IRC