2017-01-13 GnuCash IRC logs
00:20:00 *** O01eg has quit IRC
00:44:08 *** fell has quit IRC
01:27:00 *** Mechtilde has joined #gnucash
02:07:31 *** iliv has joined #gnucash
02:16:14 *** Mechtilde has quit IRC
03:00:31 *** Brent has quit IRC
03:20:48 *** Mechtilde has joined #gnucash
03:22:19 *** cartsoftware has quit IRC
03:32:00 *** fekepp has quit IRC
04:01:38 *** fekepp has joined #gnucash
04:21:12 *** meb has joined #gnucash
04:27:07 *** meb has quit IRC
04:27:56 *** meb has joined #gnucash
04:34:55 *** gjanssens has joined #gnucash
04:34:55 *** ChanServ sets mode: +o gjanssens
04:54:04 *** fabior has joined #gnucash
05:09:11 *** fabior has quit IRC
05:20:21 *** karelk has quit IRC
05:20:59 *** fabior has joined #gnucash
05:21:29 <gjanssens> .
05:21:29 <gncbot> gjanssens: Sent 10 hours and 48 minutes ago: <jralls> I tried passing bogus dates to all of the relevant tests in test-gnc-date.c and got no exceptions. I checked out your cpp branch and got no test failures at all, never mind exceptions.
05:21:30 <gncbot> gjanssens: Sent 9 hours and 47 minutes ago: <jralls> Never mind, I messed up a build parameter so I tested the wrong source dir. Now I'm getting exceptions.
05:21:31 <gncbot> gjanssens: Sent 9 hours and 23 minutes ago: <jralls> I've fixed and pushed the gnc_dmy2timespec_neutral exception leaks. It now PWARNS and returns INT64_MAX. I'll go through the rest of gnc-date.cpp and test/catch any other constructor calls tomorrow.
05:22:10 <gjanssens> jralls: I saw your commit this morning
05:23:29 <gjanssens> I didn't expect you to prevent exceptions to leak from gnc_dmy2timespec frankly.
05:23:54 <gjanssens> I thought it would just be normalized to one single exception (so rethrowing all other exceptions)
05:24:11 *** fabior has quit IRC
05:25:32 <gjanssens> But I didn't realize gnc_dmy2timespec_neutral is part of the C api...
05:25:50 <gjanssens> Is there a c++ equivalent somewhere I can use directly that does throw ?
05:26:12 <gjanssens> I found that pretty handy actually :)
05:26:22 *** fabior has joined #gnucash
05:29:35 *** fabior has quit IRC
05:42:45 *** fabior has joined #gnucash
05:55:09 *** cartsoftware has joined #gnucash
06:15:02 *** rubdos has quit IRC
06:40:29 *** rubdos has joined #gnucash
06:48:06 *** fabior has quit IRC
06:59:18 *** Jimraehl1 has joined #gnucash
07:16:10 <warlord> queso: that's exactly what you needed to do
07:16:23 *** rubdos has quit IRC
07:37:36 *** rubdos has joined #gnucash
07:57:03 *** rickoehn has joined #gnucash
08:11:20 *** fell has joined #gnucash
08:19:09 *** fabior has joined #gnucash
08:21:23 *** iliv has quit IRC
08:27:16 *** fabior has quit IRC
08:33:35 *** iliv has joined #gnucash
08:42:14 *** fabior has joined #gnucash
08:46:55 <gjanssens> The good news: travis now runs the google test based unit tests
08:47:15 <gjanssens> The bad news: several of them fail on travis while they work fine locally :(
08:48:32 <gjanssens> Note that yesterday's commit by jralls already broke the one none gtest test in libqof
08:48:53 <gjanssens> Enabling gtest now also breaks test-gnc-timezone
08:49:52 <gjanssens> jralls: any idea what could trigger this ?
08:50:33 <gjanssens> I'm suspecting the boost version being used as I'm also seeing build errors because of that in my c++ version of the csv importer
09:00:16 <gjanssens> Fyi: the travis build right after jralls' commit: https://travis-ci.org/Gnucash/gnucash/builds/191500747
09:00:40 <gjanssens> Travis build after enabling google test: https://travis-ci.org/Gnucash/gnucash/builds/191635984
09:04:21 *** fabior has quit IRC
09:04:31 <gjanssens> Regarding google test activation - I have chosen to clone google test from the primary repository on github
09:04:52 <gjanssens> By default it uses the master branch, but perhaps we'd rather use a stable release ?
09:05:36 <gjanssens> I don't know which version to choose in that case. Perhaps we should go with the oldest version we wish to support...
09:13:51 *** cartsoftware has quit IRC
09:16:49 <queso> warlord: Great, thank you. :)
09:17:03 <warlord> you're welcome
09:50:18 *** cartsoftware has joined #gnucash
09:54:12 *** karelk has joined #gnucash
10:04:39 *** queso has quit IRC
10:38:59 *** mlncn has quit IRC
10:45:09 <lmat> gjanssens: I tried to figure out what travis is using to build...I think it's using docker, so it would be nice to see his dockerFile :-) I tried to cook something up with similar versions.
10:47:37 <gjanssens> lmat: we're not using their docker instantiation but a full vm
10:48:02 <gjanssens> lmat: but even a docker file close to their environment would be helpful
10:48:23 <gjanssens> I'm currently building boost 1.54 on my Fedora 25 machine to run tests against
10:48:26 <lmat> gjanssens: Travis is running in a virtual machine?
10:48:58 <gjanssens> lmat: perhaps :)
10:49:09 <gjanssens> I meant we are using Travis-CI's virtual machine offering
10:49:21 <gjanssens> And not their docker offering
10:49:53 <gjanssens> The default (stable) docker images they provide use Ubuntu 12.12 IRC, which is too old
10:50:44 <gjanssens> I read yesterday there's a beta docker setup using 14.04LTS, but I don't know how stable that really is and don't feel like beta testing travis itself. We have enough on our plate with gnucash
10:51:48 <gjanssens> So for now gnucash sets sudo: required, which will cause Travis-CI to run our tests in a (trusty based) vm instead of a docker instance
10:53:46 <lmat> gjanssens: Good to know!
10:54:15 *** O01eg has joined #gnucash
11:00:46 <gjanssens> lmat: I remember reading in the irc logs you set up a docker instance a while back to experiment in a travis like environment
11:00:56 <gjanssens> Do you still have this available somewhere ?
11:06:49 *** fekepp has quit IRC
11:09:07 <warlord> gjanssens: is travis at all related to jenkins? Would it make sense to set up our own environment that we control completely? Or would that be a lot of work?
11:10:05 *** mlncn has joined #gnucash
11:11:37 *** Mechtilde has quit IRC
11:15:06 <lmat> gjanssens: Yes, it's online. It is best if you "know" docker before you dive in... Let me see if I can find a link.
11:16:16 <lmat> gjanssens: Let me know if this link isn't satisfactory: https://cloud.docker.com/app/lmat/repository/docker/lmat/gnucash-build/general
11:18:36 *** fekepp has joined #gnucash
11:19:14 <gjanssens> lmat: thanks
11:19:32 <gjanssens> I don't "know" docker yet, but will keep the link for later when I have more time
11:20:10 <gjanssens> warlord: I have no idea about the origin of travis, nor how much work it would be to set up our own environment
11:20:48 <gjanssens> The nice thing about travis-ci is it integrates well with github, but I'm not married to it
11:22:11 <gjanssens> Hmm, self-building boost 1.54.0 turns out to be messier than I hoped...
11:22:41 <gjanssens> My first attempt has 124 failed updating targets
11:22:51 <gjanssens> Don't know if that's bad or not
11:23:21 <gjanssens> Still the include directory is missing from the prefix dir, although the build instructions suggest it should have been created
11:23:52 *** iliv has quit IRC
11:23:53 * gjanssens thinks this investigation has to wait a bit
11:24:34 <gjanssens> I have 2 VAT declarations to finish by next thursday and there's still plenty of preparation to do for that :(
11:28:57 <warlord> good luck, gjanssens
11:34:13 *** mlncn has quit IRC
12:03:41 <gjanssens> tx warlord :)
12:12:19 *** iliv has joined #gnucash
12:13:02 *** Artefact2 has quit IRC
12:14:00 *** Artefact2 has joined #gnucash
12:21:18 *** fabior has joined #gnucash
12:23:27 *** Mechtilde has joined #gnucash
12:34:16 *** mlncn has joined #gnucash
12:39:00 *** mlncn has quit IRC
12:39:38 *** mrklintscher has joined #gnucash
12:45:12 *** fabior has quit IRC
13:00:17 *** fabior has joined #gnucash
13:36:13 *** mlncn has joined #gnucash
13:49:20 *** fabior has quit IRC
13:49:40 <jralls> gjanssens: It's too hard to debug travis directly so I have a Ubuntu 14.04 VM.
13:53:40 <jralls> gjanssens: As for GTest, +1 on using the repo. Most distros use 1.7.0, which is a tagged release in github and so easy to duplicate.
13:59:10 <jralls> gjanssens: I'd suggest that for Travis just to be consistent.
14:07:19 <jralls> gjanssens: There's no C++ equivalent to gnc_dmy2timespec_neutral. I think the right approach would be `time64 GncDateTime::neutral()`.
14:27:29 *** cyphase has quit IRC
14:42:40 *** cyphase has joined #gnucash
14:49:42 *** Mechtilde has quit IRC
15:03:20 *** Mechtilde has joined #gnucash
15:03:30 <jralls> gjanssens: Hmm, "release-1.7.0" is just gtest, no gmock. I guess we need release-1.8.0.
15:06:46 <gjanssens> jralls: ok, I'll add a checkout step in travis
15:06:59 <gjanssens> jralls: `time64 GncDateTime::neutral()` sounds good
15:08:09 * gjanssens should really invest some time in setting up a number of different development vm's as well... So much to do, so little time :(
15:08:38 <gjanssens> jralls: can you replicate the test failures in your Ubuntu 14.04 VM?
15:13:33 <warlord> gjanssens, jralls -- do you guys object if I reboot the win10 VM? I want to upgrade the ovirt infrastructure from EL7.2 to EL7.3 and ovirt 4.0.5 -> 4.0.6.
15:13:58 <gjanssens> warlord: ok with me, I'm not logged in
15:14:56 <warlord> I need to update the engine first, and then bring down all the VMs in order to update the host.
15:15:10 <warlord> but wanted to okay it with you guys first.
15:15:50 * gjanssens is signing off for today...
15:15:59 <gjanssens> See you guys later
15:16:12 <warlord> have a good night and a great weekend
15:16:37 *** iliv has quit IRC
15:18:27 <gjanssens> Thanks. Last thing I did is locking the google test version to 1.8.0, just pushed to master
15:18:37 <gjanssens> Have a nice weekend too!
15:19:34 *** gjanssens is now known as gjanssens_afk
15:22:17 *** gjanssens_afk has quit IRC
15:24:13 *** kael has joined #gnucash
15:29:59 <jralls> ls
15:30:25 <jralls> Sigh. Warlord, I'm not using the VM either, go ahead and reboot.
15:30:51 <warlord> okay. thanks.
15:31:10 <warlord> I'm working on the engine vm first. Once that's upgraded I'll work on bringing down everything else and updating the host.
15:32:02 <jralls> @tell gjanssens After fixing the test-qof failures and locking gtest all tests pass on my Ubuntu VM.
15:32:02 <gncbot> jralls: The operation succeeded.
15:32:20 <jralls> warlord: Engine VM?
15:33:38 <warlord> ovirt engine. part of the ovirt system
15:34:04 <jralls> OIC.
15:40:28 <jralls> Dang. Failed on Travis.
15:41:21 <warlord> :(
15:47:25 *** kael has quit IRC
16:14:20 *** warlord has quit IRC
16:28:15 *** fabior has joined #gnucash
16:31:21 *** fabior has quit IRC
16:34:36 *** fabior has joined #gnucash
17:00:36 *** gncbot sets mode: +o fell
17:07:58 *** fabior has quit IRC
17:11:02 *** fabior has joined #gnucash
17:15:25 *** Nytram has joined #gnucash
17:16:23 *** fabior has quit IRC
17:20:22 *** warlord has joined #gnucash
17:20:22 *** gncbot sets mode: +o warlord
17:21:40 *** iliv has joined #gnucash
17:35:37 *** iliv has quit IRC
17:55:33 *** meb has quit IRC
17:58:51 *** meb has joined #gnucash
18:09:36 *** rickoehn has quit IRC
18:25:59 <jralls> @tell gjanssens: Travis tests fixed. The TimezoneProvider constructor didn't actually check the TZ variable if it was passed a bogus zone file. Now it does.
18:25:59 <gncbot> jralls: The operation succeeded.
18:27:08 *** Mechtilde has quit IRC
18:27:44 *** movedx has joined #gnucash
18:28:09 <movedx> Hello all :)
18:42:15 *** storyjesse has joined #gnucash
19:20:26 *** Mechtilde has joined #gnucash
19:29:19 *** Mechtilde has quit IRC
20:02:59 *** Steve has joined #gnucash
20:21:24 *** mlncn has quit IRC
20:44:33 *** edgardleal has joined #gnucash
20:45:48 *** cyphase has quit IRC
20:49:57 *** linas has joined #gnucash
20:51:46 *** edgardleal has quit IRC
21:00:40 *** Nytram has quit IRC
21:59:42 *** User_ has joined #gnucash
22:10:34 *** mlncn has joined #gnucash
23:03:19 *** User_ has quit IRC