2018-01-17 GnuCash IRC logs
00:58:21 *** harshitaneja has joined #gnucash
01:00:41 *** jotrago has quit IRC
01:29:51 *** marusich has joined #gnucash
01:29:53 *** omnireq has joined #gnucash
01:33:55 *** Mechtilde has joined #gnucash
01:37:00 *** fell_afk has quit IRC
01:38:08 *** fell_afk has joined #gnucash
02:03:52 *** gjanssens has joined #gnucash
02:03:52 *** ChanServ sets mode: +o gjanssens
02:04:09 <gjanssens> .
02:05:59 <Mechtilde> Good morning
02:06:14 <Mechtilde> gjanssens, thanks for merging the PR
02:12:22 *** fabior has joined #gnucash
02:14:41 <gjanssens> Mechtilde: good morning and you're welcome
02:16:18 *** Mechtilde has quit IRC
02:49:29 *** fekepp has quit IRC
03:36:18 *** bastianilso_ has quit IRC
03:36:43 *** Mechtilde has joined #gnucash
03:45:02 *** Cork has quit IRC
03:45:02 *** fabior has quit IRC
03:45:02 *** fell_afk has quit IRC
03:45:02 *** marusich has quit IRC
03:45:02 *** omnireq has quit IRC
03:45:02 *** GabrieleV has quit IRC
03:45:02 *** jchonig has quit IRC
03:45:46 *** fabior has joined #gnucash
03:45:46 *** fell_afk has joined #gnucash
03:45:46 *** omnireq has joined #gnucash
03:45:46 *** marusich has joined #gnucash
03:45:46 *** GabrieleV has joined #gnucash
03:45:46 *** jchonig has joined #gnucash
03:49:39 *** pilotauto has quit IRC
03:50:23 *** Cork has joined #gnucash
03:53:35 *** fell_afk is now known as fell
03:53:58 *** gncbot sets mode: +o fell
04:03:08 *** bastianilso_ has joined #gnucash
04:05:29 *** bastianilso_ has quit IRC
04:08:03 *** bastianilso_ has joined #gnucash
04:14:31 *** bertbob has quit IRC
04:39:52 <fell> I am wondering why travis fails on maint with null log. A local 'make check' succeeds.
04:40:44 *** fabior has quit IRC
04:44:41 *** jotrago has joined #gnucash
04:48:27 *** bertbob has joined #gnucash
04:53:49 *** jotrago has quit IRC
05:13:11 *** Aussie_matt has joined #gnucash
05:23:19 *** fekepp has joined #gnucash
05:23:30 *** bastianilso_ has quit IRC
05:36:11 *** jotrago has joined #gnucash
05:55:10 *** jotrago has quit IRC
05:55:46 *** jotrago has joined #gnucash
06:06:21 *** jotrago1 has joined #gnucash
06:06:48 *** jotrago has quit IRC
06:06:49 *** jotrago1 is now known as jotrago
06:17:12 *** jotrago1 has joined #gnucash
06:17:40 *** jotrago has quit IRC
06:17:41 *** jotrago1 is now known as jotrago
06:30:34 *** storyjesse has joined #gnucash
06:56:17 *** Jimraehl1 has joined #gnucash
06:59:30 *** Jimraehl1 has quit IRC
07:24:01 *** bob_ has joined #gnucash
07:46:50 *** jotrago has quit IRC
07:46:55 *** jotrago has joined #gnucash
07:57:04 *** bob_ has quit IRC
08:03:23 *** User has joined #gnucash
08:08:02 *** Aussie_matt has quit IRC
08:16:54 *** User has quit IRC
08:17:25 *** oozer has joined #gnucash
08:26:10 *** chris has quit IRC
08:27:15 *** marusich has quit IRC
08:27:35 *** marusich has joined #gnucash
08:29:23 <warlord> @op
08:29:23 *** gncbot sets mode: +o warlord
08:38:13 *** omnireq has quit IRC
08:45:21 <warlord> .
08:49:57 *** chris_ has joined #gnucash
08:55:37 *** jotrago has quit IRC
09:06:53 *** Mechtilde has quit IRC
09:19:16 *** fabior has joined #gnucash
10:13:23 *** kael has joined #gnucash
10:13:57 *** User has joined #gnucash
10:20:02 <gjanssens> fell: the travis failure is probably a temporary thing... My last commit passed all travis tests and was a single string that doesn't affect any tests.
10:28:34 *** User has quit IRC
10:30:41 <gjanssens> fell: I'm trying to merge maint into unstable, but I'm getting a massive merge conflict on de.po (655 unresolved merge conflicts)
10:31:37 <gjanssens> Can I just ignore the po file on maint and recreate the po file on unstable after merging your translation comment fixes ?
10:32:11 <lmat> jralls_afk: "Do you have an arch installation that duplicates..." Are you talking about https://travis-ci.org/Gnucash/gnucash/jobs/328524662 ? Sorry, I've been out of band for a while now.
10:32:11 <fell> gjanssens, yeah, but I was more confused by the other conflicts.
10:33:12 <gjanssens> The others *I* can understand and handle :)
10:34:52 <fell> OTOH a rebase, I worked fine. So I was thinking about saving a locally rebased de.po and copy it in the resolving
10:36:44 <gjanssens> What did you rebase on what ?
10:38:41 <fell> I did in my local unstable branch a git merge maint and de.po had no conflicts.
10:39:12 <gjanssens> fell: why did you split the (define-report ...) lines in two in a number of standard-reports ? In commit 762c3b8f56939201bb72c55125d292cc645809d8
10:39:32 <fell> Stop, I didn't read caoutious enough.
10:39:34 <gjanssens> fell: is your unstable branch fully up to date with our github repo ?
10:39:55 <fell> It was last night
10:40:21 <fell> To suppress unwanted comments in gnucash.pot
10:41:09 <fell> xgettext seems to look backward from _()
10:42:05 <fell> If the last not empty line contains a comment, it will copy it.
10:42:22 <gjanssens> Oh, I see.
10:43:01 <gjanssens> In this case I would have removed the commented out text, but this is fine as well. I'll reproduce your changes on unstable.
10:45:13 <lmat> jralls_afk: I just built bye_bye_autotools and it succeeded. I'll try updating my machine and testing again.
10:52:08 <fell> So, I remember, when I did the rebase, the only conflicts were in engine.c and it's test
10:53:07 <fell> engine/Account.c
10:56:39 *** ArtGravity has joined #gnucash
10:58:08 <gjanssens> fell: Ok, but that doesn't make sense to me. In the best case the rebase may have less conflicts, but the will resurface when you attempt to merge maint aferwards. They're not gone, you may have temporarily elided them only.
10:58:52 <gjanssens> And I wonder whether your unstable branch was fully up to date with current unstable. Did you pull origin on unstable before attempting the rebase ?
10:59:30 <fell> Looking back, I should have done a merge after changing the sources and then run make pot;msgmerge on unstable before continuing with the rest.
11:01:42 <gjanssens> Perhaps, but I suspect we will continue to see po file merge issues for as long as we accept changes on both unstable and master
11:02:38 <gjanssens> Contrary to source files the changes in these files are much more disruptive. The msgmerge tool is not very vcs friendly it seems.
11:02:43 <fell> That's the reason I did't apply the PR
11:03:10 <gjanssens> Mechtilde's PR you mean ?
11:03:16 <fell> Yep
11:03:34 <gjanssens> To unstable?
11:03:38 <fell> I wanted apply it after my review on maint
11:03:51 <gjanssens> I wouldn't have mattered
11:04:11 <gjanssens> The disruption started when I imported the goffice translatable strings :(
11:05:03 <fell> Perhaps we shoud revert and reorder the commits?
11:05:12 <gjanssens> Mechtilde's PR was heavily impacted by that as well, and I had to handle about 150 unresolvable merge conflicts
11:05:59 <gjanssens> Again reordering the commits wouldn't change things, except I'd have to handle the merge conflicts once more
11:06:32 <gjanssens> It's a call to xgettext on the unstable branch that starts the conflict.
11:06:35 *** kael has quit IRC
11:07:02 <gjanssens> As we announced 2.7.3 to be the string freeze that was the normal thing to do.
11:07:31 <fell> I did not change any strings
11:07:50 <gjanssens> In maint you mean ?
11:07:56 <fell> I removed one never shown and changed comments
11:08:03 <fell> yes
11:08:11 *** storyjesse has quit IRC
11:09:47 <gjanssens> That's fine. The core issue is though that xgettext reorders strings so a po file merged with the unstable gnucash.pot file is in essence unmergable with a po file merged with the maint gnucash.pot
11:09:56 <gjanssens> Regardless of any other changes made to the file
11:10:54 <fell> The order is by 1. occurrence in dir/file
11:11:15 <fell> But we changed also the directory structure
11:12:17 <fell> While git can handle the change, if it's buffer is big enough, xgettext is agnostic here
11:14:21 *** fabior has quit IRC
11:16:22 <gjanssens> fell: That's probably exactly the issue indeed.
11:16:27 <fell> commit 3c1dc14 was the head of my unstable branch
11:17:07 <gjanssens> Due to the changed directory structure, the messages got reordered as well. So in this case it's a massive file change.
11:17:48 <fell> and a normal diff/patch will fail
11:17:59 <gjanssens> The same could have happened in other situations on a smaller scale by renaming files/moving them to other directories/deleting files/adding files
11:18:36 <gjanssens> At some point the differences are too far from one another to spot their relation as a human.
11:19:03 <gjanssens> Git will resolve anything that doesn't conflict correctly, it gets hard when a human as to look at the unresolved conflicts.
11:19:22 <gjanssens> Oh, and git doesn't even handle non-conflicts correctly!
11:20:29 <gjanssens> I had to restart the merge of Mechtilde's PR because it had incorrectly omitted an empty msgstr "" and added a msgstr twice in a few other locations. Those were far away from any unresolvable conflicts
11:20:39 <fell> For git, a line is a record, for xgettext, a section
11:21:05 <gjanssens> Bottom line is, there was a reason we had a policy that translation work should only happen on one branch at the time :)
11:21:32 <gjanssens> We should update our translation status page to inform translators they should now use unstable rather than maint.
11:22:18 <gjanssens> I was working on that :)
11:23:20 <fell> So you do not allow fixes for 2.6.20?
11:27:57 *** fabior has joined #gnucash
11:28:39 <fell> IIRC cstim had 2 alternate texts in the status page and commented out the not current version.
11:28:51 <gjanssens> fell: I don't know
11:29:05 <gjanssens> (about fixes for 2.6.20)
11:29:40 <gjanssens> The alternate text seems to have gotten messed up at some point. What's there is not the correct message for unstable work.
11:30:23 <gjanssens> About 2.6.20, we can accept fixes, but we risk losing them on unstable due to the merge conflicts.
11:30:50 <gjanssens> So possibly the translator has to do the work twice to get the fixes in 3.0 as well eventually.
11:31:11 <gjanssens> I don't know which option is the least bad one...
11:33:50 <gjanssens> I'll do some experiments with this later to estimate how bad it would be
11:35:31 *** harshitaneja has quit IRC
11:37:14 <gjanssens> https://wiki.gnucash.org/wiki/Translation_Status
11:37:20 <gjanssens> fell: ^
11:37:48 <fell> I wonder, if some SQL wizzardry could do the resolve for us:
11:39:09 <fell> gettext records have a primary key MsgID, but the sort order in file is by first occurrence in directory/file:line
11:40:18 <gjanssens> fell: I have been pondiering on and off about an online translation system (self-hosted or not)
11:40:32 <gjanssens> I believe pootle was an interesting candidate
11:41:06 <gjanssens> And I believe such a system could greatly simplify the lives of translators.
11:41:25 <gjanssens> But I have no time right now to go into this too deeply
11:41:35 <gjanssens> I have some real accounting to finish by Friday...
11:45:34 <fell> I remember one pt_BR.po patch in BZ ...
11:45:45 <fell> It is trivial
11:46:08 <fell> but I am not happy with the content
11:46:49 <fell> The replace "Process payment" by retranslated "Process Invoice"
11:47:05 <fell> They
11:48:47 <gjanssens> Yes, I've seen that one as well
11:49:20 <gjanssens> It's a poorly chosen translation as we have Process payment for invoices, bills and employee vouchers
11:49:31 <gjanssens> It's not alwas an invoice that gets paid
11:49:57 * gjanssens should add his feedback there
11:56:00 *** marusich has quit IRC
11:57:29 *** fekepp has quit IRC
12:06:47 <gjanssens> Done
12:07:12 * gjanssens tends to do lots of other things to postpone accounting :(
12:08:36 <gjanssens> fell: I have pushed the merge from maint to unstable
12:09:09 <fell> :)
12:09:22 *** fabior has quit IRC
12:09:48 <fell> Not only gjanssens ;-)
12:10:30 <fell> But the tax office will not wait ...
12:11:44 <gjanssens> Nope
12:12:07 <gjanssens> Unless they have to pay :D
12:56:39 *** gnomey has quit IRC
12:58:12 *** gnomey has joined #gnucash
13:13:26 *** fabior has joined #gnucash
13:31:29 *** fabior has quit IRC
14:04:02 *** fekepp has joined #gnucash
14:06:19 *** frakturfreak has joined #gnucash
14:18:06 *** Mechtilde has joined #gnucash
14:27:42 *** fabior has joined #gnucash
14:30:59 *** frakturfreak has quit IRC
14:32:19 *** frakturfreak has joined #gnucash
14:50:27 *** gjanssens has quit IRC
14:52:03 *** gjanssens has joined #gnucash
14:52:03 *** ChanServ sets mode: +o gjanssens
15:23:48 *** Mechtilde has quit IRC
15:25:19 *** Mechtilde has joined #gnucash
15:25:58 *** frakturfreak has quit IRC
15:27:03 *** frakturfreak has joined #gnucash
16:00:37 *** Mechtilde has quit IRC
16:42:15 *** frakturfreak has quit IRC
16:46:18 *** fabior has quit IRC
17:04:29 *** gjanssens has quit IRC
17:54:56 *** fell is now known as fell_afk
18:15:26 *** fekepp1 has joined #gnucash
18:16:02 *** fekepp has quit IRC
18:58:45 *** pilotauto has joined #gnucash
19:29:15 *** ArtGravity has quit IRC
22:04:16 *** kael has joined #gnucash
22:09:49 *** oozer has quit IRC
22:24:54 *** kael has quit IRC
22:56:45 *** meb has quit IRC
22:58:42 *** meb has joined #gnucash
23:02:28 *** meb has quit IRC