2017-08-21 GnuCash IRC logs

00:02:04 *** O01eg has joined #gnucash
00:21:06 *** storyjesse has quit IRC
00:49:32 *** Mechtilde has joined #gnucash
01:18:49 *** storyjesse has joined #gnucash
01:24:53 *** fell has quit IRC
01:25:42 *** fell has joined #gnucash
01:26:36 *** bhardwajs has quit IRC
01:30:57 *** Mechtilde has quit IRC
02:19:39 *** cyphase has quit IRC
02:20:39 *** cyphase has joined #gnucash
02:33:11 *** cyphase has quit IRC
02:38:38 *** cyphase has joined #gnucash
02:44:54 *** jotrago1 has joined #gnucash
02:45:48 *** jotrago has quit IRC
02:45:49 *** jotrago1 is now known as jotrago
03:11:59 *** mrklintscher16 has joined #gnucash
03:13:27 *** mrklintscher15 has quit IRC
03:15:08 *** mrklintscher16 has quit IRC
03:31:17 *** mrklintscher16 has joined #gnucash
03:39:10 *** mrklintscher16 has quit IRC
03:40:39 *** Mechtilde has joined #gnucash
03:41:04 *** Riviera has quit IRC
03:42:12 *** storyjesse1 has joined #gnucash
03:52:06 *** fabior has joined #gnucash
03:54:22 *** mrklintscher16 has joined #gnucash
04:24:46 *** gorgabal has joined #gnucash
04:57:25 *** gjanssens has joined #gnucash
04:57:25 *** ChanServ sets mode: +o gjanssens
04:57:48 <gjanssens> .
05:01:00 *** jotrago has quit IRC
05:02:22 *** gorgabal has quit IRC
06:11:36 *** storyjesse has quit IRC
06:11:54 *** storyjesse1 is now known as storyjesse
06:17:11 *** bhardwajs has joined #gnucash
06:19:49 *** pilotauto has quit IRC
06:50:42 *** fabior has quit IRC
07:10:37 *** gorgabal has joined #gnucash
07:47:52 *** rickoehn has joined #gnucash
07:58:08 *** gorgabal has quit IRC
08:00:45 *** Mechtilde has quit IRC
08:00:59 *** Mechtilde has joined #gnucash
08:07:35 *** Mechtilde has quit IRC
08:24:12 *** Mechtilde has joined #gnucash
08:25:57 *** mib_achu21 has joined #gnucash
08:26:27 *** new_to_gnucash has joined #gnucash
08:26:37 *** mib_achu21 has quit IRC
08:26:41 <new_to_gnucash> Hi all
08:26:57 <new_to_gnucash> Can someone give a quick hand with GnuCash?
08:29:48 <new_to_gnucash> I cant post my bill to my A/P acc
08:29:54 <new_to_gnucash> The field is all messed up - https://i.imgur.com/ZjY7iYd.png
08:30:11 <new_to_gnucash> Capture - https://i.imgur.com/ZjY7iYd.png
08:30:27 <new_to_gnucash> Can I do anything about it?
08:34:30 <new_to_gnucash> Same goes for Invoices to A/R :(
09:00:34 *** kael has joined #gnucash
09:08:05 *** fabior has joined #gnucash
09:09:23 *** gncbot sets mode: +o fell
09:10:41 <fell> new_to_gnucash: You need at least one A/R account in the same currency as the bill has.
09:13:49 <fell> So check, which is your default currency, which have your existing A/R accounts, which the bill.
09:14:20 <new_to_gnucash> Makes sense
09:14:28 <new_to_gnucash> I guess that would be it
09:15:23 *** jralls_afk is now known as jralls
09:16:17 <new_to_gnucash> Yep. Got it. :)
09:16:19 <new_to_gnucash> Thank yoU!
09:17:03 <fell> :-)
09:20:47 *** jotrago has joined #gnucash
09:32:59 <fell> gjanssens: What is the best practice, if a PR author used the maint branch?
09:33:11 *** new_to_gnucash has left #gnucash
09:33:45 <fell> Should I ammend the commit to set me as committer or should I pullt in a new branch end the merge?
09:36:02 <gjanssens> fell: good question
09:37:10 <gjanssens> What do you want to achieve?
09:37:34 <gjanssens> I think you can safely pull the remote maint branch into yours.
09:38:08 <gjanssens> It will either be a fast forward (if it was based on your current maint head) or a merge operation merging a remote maint branch into your local one.
09:38:54 <fell> I can now test option 1 and if that is not recognized by github, we can use option 2 in the future.
09:38:57 <gjanssens> In the case of a merge the merge comment will mention <remote>/maint as origin branch, which is fine IMO
09:39:11 <gjanssens> It will be recognied by github.
09:39:42 <gjanssens> Github will only not recognize it if the commit's hashes change, which happens only when you rebase or cherry-pick
09:40:11 <gjanssens> after pulling the branch locally.
09:40:44 <gjanssens> I'm doing this less and less to stick with git's and github's original intents
09:52:57 <fell> 'git pull https://github.com/valdisvi/gnucash maint' did not work. It had to be 'git pull https://github.com/valdisvi/Gnucash maint'
09:56:43 <gjanssens> fell: that would surprise me. I tried both and both links exists (and are identical by the way - github seems to ignore case in repo names)
09:57:11 <gjanssens> Are you sure there wasn't another typo ?
09:57:33 <fell> It should, but it didn't.
09:57:53 <gjanssens> Ok. *shrugs*
09:57:54 <fell> It was the only letter which I changed
09:58:05 <gjanssens> Perhaps a temporary error...
09:59:37 <gjanssens> I seem to be having some issues pulling from github right now
10:08:03 <gjanssens> fell: I think this is something with github atm. Trying a couple of pulls with exactly the same remote url, some will work others will time out
10:08:24 <gjanssens> not much we can do about this but waiting for github to fix it...
10:08:29 <fell> ACK
10:15:29 <fell> gjanssens, my pull resulted in
10:15:30 <fell> Zunächst wird der Branch zurückgespult, um Ihre Änderungen
10:15:32 <fell> darauf neu anzuwenden...
10:15:33 <fell> Wende an: Partial fix for bug 660027
10:15:35 <fell> Wende an: Remove clutter from generated html
10:15:50 <fell> Rewind, apply
10:17:40 <gjanssens> fell: it looks like your local maint branch has diverged from origin's maint branch
10:18:00 <gjanssens> Did you have local commits on maint that weren't pushed to code yet ?
10:18:27 <fell> No, the lv patch was older than the 2 html patches
10:18:40 <gjanssens> (gitk is a helpful graphical tool to visualize the state of your repo's branches)
10:18:57 <gjanssens> Ok, so I presume you ended up with a merge?
10:19:11 <gjanssens> A merge *commit* I meant ?
10:19:18 <fell> No conflict, no merge
10:20:24 <fell> but it behaves like a merge branch.
10:20:46 <gjanssens> Hmm, I'm having a hard time seeing the full picture. Can you post a screenshot from the window that opens with the command "gitk --all" ?
10:21:54 <gjanssens> fell: I'm mostly interested in the upper half of the window, the part showing the commit list
10:22:35 <gjanssens> Oh wait: what was your complete pull command? Did you add --rebase ?
10:23:07 <fell> I would have to send 2 pages because of the changes in master
10:23:40 <gjanssens> Oh, right. Just drop the '--all' part then. That should show enough also
10:30:18 <fell> with "maint remotes/origin/maint" I got the right
10:33:52 <gjanssens> fell: the right what ?
10:34:04 <fell> selection
10:34:16 <gjanssens> For a screenshot ?
10:34:23 <fell> yes
10:34:27 <gjanssens> Ok
10:46:41 <fell> gjanssens: https://imgur.com/a/dRF1u
10:48:36 <gjanssens> fell: are you using git pull --rebase ?
10:49:35 <gjanssens> If so, I think you shouldn't because it's doing the rebase backwards
10:49:36 <fell> yes
10:50:04 <gjanssens> What it does is it pulls in the remote branch and then rebases your local branch on top of that
10:50:42 <fell> No rebase, when pulling lv
10:51:13 <gjanssens> Your local branch was your local maint branch which had two commits that were not on the remote branch (yet)
10:51:23 <gjanssens> What do you mean with "pulling lv" ?
10:51:58 <gjanssens> The previous pull from the valdisvi?
10:53:05 <fell> I was on maint:1559d26
10:53:31 <fell> then entered: git pull https://github.com/valdisvi/Gnucash maint
10:53:53 <fell> and got above messages
10:54:20 <gjanssens> git pull *--rebase* https://github.com/valdisvi/Gnucash maint I presume
10:54:21 <fell> Rewinding and reapplying the 2 last commits
10:55:27 <gjanssens> These messages tell you git rebased your maint on top of Valdis' maint. Which is backwards and the direct result of using --rebase
10:55:27 <fell> above is the copy from my shell, so no explicit --rebase
10:55:43 <gjanssens> Then there must be an implicit rebase going on on your system
10:56:01 <gjanssens> Did you add something to that effect to your git config file ?
10:56:43 <fell> Hu, where do I find it?
10:59:14 <gjanssens> fell: what's in .git/config in your repository ?
11:01:45 <gjanssens> fell: and what's the output of git config pull.rebase
11:02:27 <fell> error: key does not contain a section: pull
11:05:11 <gjanssens> It's not "pull" it's "pull.rebase"
11:05:33 <fell> New line
11:06:01 <fell> empty
11:09:29 <gjanssens> Ok, you are running this inside your gnucash repo, right ?
11:09:37 <gjanssens> Here are two others to test:
11:09:46 <fell> I yes
11:09:58 <gjanssens> git config branch.autoSetupRebase
11:10:08 <gjanssens> git config branch.maint.rebase
11:10:34 <fell> git config branch.autoSetupRebase: empty
11:11:14 <fell> git config branch.maint.rebase: true
11:11:38 <gjanssens> Bingo. That setting is causing your pull requests to automatically perform a rebase instead of a merge
11:13:09 <fell> That is in .git/config set for both branches
11:13:41 *** mrklintscher16 has quit IRC
11:13:42 <gjanssens> You should remove it. It doesn't do what you'd expect
11:14:16 <gjanssens> As I tried to explain above it will perform a backwards rebase if a rebase is needed.
11:14:29 <fell> direct in .git/config or with git config something
11:14:37 <gjanssens> Both are fine
11:14:43 <gjanssens> direct edit is probably the easiest
11:15:28 <gjanssens> When you remove these options pull requests can result in a merge from the remote branch into your locally active one.
11:16:06 <gjanssens> I used to manually undo the merge and then rebase the remote branch onto my local branch, but I think it's time we stop doing that.
11:16:20 <fell> I don't remember, from where it came
11:16:37 <gjanssens> It makes things needlessly complicated for all parties.
11:16:59 <gjanssens> It's possible I suggested something to that effect to you in the past. If so that was a mistake.
11:17:26 <fell> Can you add the strategy to https://wiki.gnucash.org/wiki/Git#Patches_and_Pull_Requests ?
11:19:04 <fell> BTW I added your recent mail as #Advanced_Pulling
11:20:28 <fell> And there you made a Sidenote already
11:20:48 <gjanssens> fell: thinking about this some more I'm afraid it's not that straightforward :(
11:20:56 <gjanssens> It depends on the context.
11:21:25 <fell> If A then if B then, ...
11:21:42 <gjanssens> A "git pull --rebase" is a useful command for those who insist on working on the two primary branches (maint/master)
11:22:18 <gjanssens> and if the user wants to pull in the latest changes on that branch without loosing his local work.
11:22:38 <fell> It is useful for pulling origin, but bad for others
11:22:48 <gjanssens> That's one more good reason to never work on one of these primary branches unless you know what you are doing.
11:22:58 <gjanssens> Heh, that's a pretty good summary :)
11:23:23 <gjanssens> Thanks
11:24:34 <fell> Is it posible to add it somehow to origin
11:28:46 <gjanssens> I don't know, I can't find that in "git help config"
11:29:16 <gjanssens> fell to proceed with your PR handling, you should first hard reset your local maint branch on origin/maint
11:29:28 <gjanssens> Then run git pull https://.... again
11:29:54 <gjanssens> If you removed the two rebase options it should now do a merge instead
11:30:21 * fell will do so ...
11:30:40 <gjanssens> Also here's how I typically handle a pull from master/maint if I still have local commits:
11:30:52 <gjanssens> git pull master (assuming I'm doing master here)
11:31:26 <gjanssens> If this command results in a merge I will reset master to where I was locally (so the commit is was on right before the merge)
11:31:40 <gjanssens> And then do git pull --rebase master
11:32:18 <gjanssens> It requires to pay attention to the output of the rebase command (which could also fail in case of merge conflicts btw)
11:32:48 <gjanssens> Yet another strategy I have used as well is
11:32:50 <gjanssens> git rebase master
11:33:11 <gjanssens> If this merges or conflicts instead of fast forwarding,
11:33:21 <gjanssens> git reset --hard origin/master
11:33:43 <gjanssens> git cherry-pick <all commits I had locally on master before pulling>
11:34:31 <gjanssens> I usually perform that last step via gitk which I have open all the time so I have a visual representation of what's going on
11:36:49 <fell> while pulling I get Merge branch 'maint' of https://github.com/valdisvi/Gnucash into maint; Enter a commit decription...
11:37:23 <gjanssens> That's normal. A 'merge' is a new commit. Just keep the message as is. It says it all...
11:38:09 <fell> I will enter PR #172: Accelerator fixes for Latvian language
11:38:30 <gjanssens> I meant to say ... A 'merge' is a new commit *so it needs a commit message*
11:38:57 <gjanssens> That's ok, but keep the merge standard message as well.
11:53:43 <fell> Next time I will pull in a new branch like lv in this case. Then I can add in the merge commit statistical data like 4685 Translated messages
11:58:54 <gjanssens> You can still do that now
11:59:03 <gjanssens> Just run git commit --amend
11:59:57 <gjanssens> Pulling in a new branch has other disadvantages
12:03:45 <gjanssens> In very short: you either lose the merge comment or you'd have to do two merge commits
12:08:32 <fell> OK, next time...
12:08:36 <fell> Ah, the eclipse push dialog has a Section "Configure upstream for push and pull"
12:15:30 <gjanssens> Next time indeed... I think you do know by now you should never alter a commit already pushed to code.gnucash.org in any way (no rebase, not amendements,...)
12:15:47 <gjanssens> (Just stating the obvious)
12:17:19 *** kael has quit IRC
12:18:59 <fell> I aborted the push dialog and applied it before the real push.
12:28:07 *** Mechtilde has quit IRC
12:45:25 *** jralls is now known as jralls_afk
12:54:24 *** kael has joined #gnucash
13:13:56 *** bhardwajs has quit IRC
13:38:37 *** Mechtilde has joined #gnucash
13:56:26 *** fabior has quit IRC
14:10:48 *** fabior has joined #gnucash
14:14:04 *** fabior has quit IRC
14:16:56 *** cyphase has quit IRC
14:17:14 *** cyphase has joined #gnucash
14:21:34 <fell> gjanssens: https://bugzilla.gnome.org/attachment.cgi?id=357993&action=diff is pretty simple in maint, but will produce a merge conflict, because you removed src on master.
14:26:08 <fell> And in https://bugzilla.gnome.org/show_bug.cgi?id=784284 is a GTK2 patch, where I don't know if it should be applied. Translators would it.
14:26:40 <fell> like
14:38:38 *** frakturfreak has joined #gnucash
14:48:02 *** Mechtilde has quit IRC
14:53:02 *** kael has quit IRC
15:00:19 *** aqua has joined #gnucash
15:13:12 <gjanssens> fell: for the first bug, did you try to merge from maint to master or did you simply try to apply the patch on master ?
15:14:43 <fell> I didn't try, only fear
15:15:47 <fell> But I can try ...
15:28:50 <gjanssens> fell: a merge has the best chances of working because git knows the original file location and usually handles renames and moves pretty well
15:29:12 <gjanssens> Simply applying a patch will probably not work because this same context is missing
15:35:11 <fell> I get 2 conflicts, 1 by lv.po and 1 by this patch. Let me see ...
15:39:10 <fell> and in the reports too
15:39:58 <gjanssens> I pulled maint and master and see several merge conflicts as well
15:40:22 <gjanssens> most of them can be solved by adding a small extra section in .git/config:
15:40:36 <gjanssens> [merge]
15:40:37 <gjanssens> renamelimit = 3000
15:41:03 <gjanssens> That reduces the merge conflict to one file: gnucash/report/report-system/html-document.scm
15:41:17 *** aqua has quit IRC
15:41:46 <gjanssens> There are two conflicts in that file that should be fairly trivial to solve.
15:41:51 <gjanssens> If you want I can do that for you.
15:42:20 <gjanssens> I haven't applied the patch from the first bug yet but I expect that to work fine as well.
15:43:55 <fell> OK, go for it. I will reset, and pull after your change.
15:44:13 *** bertbob has quit IRC
15:45:35 <gjanssens> Ok
15:56:35 <gjanssens> fell: applied, merged and pushed (veni, vidi, vici :)
15:57:09 <fell> alea iacta est ;-)
15:59:08 *** bertbob has joined #gnucash
16:00:01 <gjanssens> :D
16:02:26 *** bertbob has quit IRC
16:12:32 <fell> gjanssens: I just marked the patch as commited to get it out of list.
16:16:28 *** jralls_afk is now known as jralls
16:16:45 *** bertbob has joined #gnucash
16:17:32 <gjanssens> fell: tx
16:17:49 *** frakturfreak has quit IRC
16:18:05 <jralls> fell: Thanks for going through bugs and reassigning them to one of the psuedo addresses. It doesn't really matter if the old assignee is active or not, it's better to have the pseudo address so that we all get bugmail.
16:19:31 <fell> jralls, there are still many pending. I will continue on my next duplicate serach.
16:19:48 <jralls> Please do!
16:27:35 *** Cuare has joined #gnucash
16:30:48 <fell> gjanssens: Remains the gtk2 patch. the line numbers are very different in gtk3
16:43:14 <gjanssens> fell: I did another merge via git. It handled the line number hurdle but I still had to resolve some merge conflicts because gtk3 had obsoleted some xml tags.
16:44:24 <fell> So you are almost ready with it?
16:46:36 <fell> I see it on github. Thx!
16:47:09 <fell> But you got a yellow Point there.
16:47:28 <fell> Travis is pending
16:48:11 <gjanssens> That means it's still running and can't tell yet whether its tests have run successfully
16:48:33 <gjanssens> Travis takes between 10-20 minutes to complete.
17:13:58 <gjanssens> fell: there you go. Travis is happy
17:14:10 *** Artefact2 has joined #gnucash
17:17:16 *** jralls is now known as jralls_afk
17:53:52 *** rickoehn has quit IRC
17:58:01 *** gjanssens has quit IRC
17:59:05 *** mrklintscher16 has joined #gnucash
18:06:04 *** mrklintscher17 has joined #gnucash
18:08:25 *** mrklintscher16 has quit IRC
18:33:04 *** mrklintscher17 has quit IRC
18:38:10 *** pilotauto has joined #gnucash
19:06:18 *** bhardwajs has joined #gnucash
21:09:51 *** cyphase has quit IRC
21:15:05 *** cyphase has joined #gnucash
21:30:51 *** fell is now known as fell_afk
21:50:15 *** bertbob has quit IRC
21:51:53 *** meb has quit IRC
21:56:40 *** meb has joined #gnucash
22:05:48 *** bertbob has joined #gnucash
22:23:07 <lmat> jralls_afk: regarding "imap guid conversion down to a few seconds..." Excellent work!
22:23:21 <lmat> Quite an eclipse today.
23:46:38 *** bhardwajs has quit IRC
23:48:43 *** bhardwajs has joined #gnucash