2018-03-29 GnuCash IRC logs

00:04:43 *** fell has joined #gnucash
00:12:59 *** bertbob has quit IRC
00:14:00 *** bertbob has joined #gnucash
01:09:25 *** fell has quit IRC
01:09:41 *** Mechtilde has joined #gnucash
01:10:36 *** fell has joined #gnucash
01:43:57 *** CrashCart has quit IRC
01:46:59 *** Mechtilde has quit IRC
01:49:18 *** jotrago has joined #gnucash
01:50:57 <gjanssens_afk> fell: good tip, thanks
01:51:03 *** gjanssens_afk is now known as gjanssens
01:55:56 *** gjanssens has quit IRC
02:09:44 *** gjanssens has joined #gnucash
02:09:44 *** ChanServ sets mode: +o gjanssens
02:25:08 *** fekepp has quit IRC
03:11:43 *** fabior has joined #gnucash
03:38:36 *** fekepp has joined #gnucash
04:02:34 *** jotrago has quit IRC
04:06:30 *** jotrago has joined #gnucash
04:11:41 *** jotrago has quit IRC
04:15:45 *** ncv has joined #gnucash
04:52:10 *** jotrago has joined #gnucash
05:02:09 *** chris has joined #gnucash
05:02:50 *** pilotauto has quit IRC
05:38:27 *** jotrago has quit IRC
05:38:36 *** jotrago has joined #gnucash
05:44:30 *** weasel has quit IRC
06:03:27 *** chris has quit IRC
06:18:27 *** fabior has quit IRC
06:19:58 *** PaulW2U has joined #gnucash
06:35:44 *** jiangfuqiao has joined #gnucash
06:48:34 *** fabior has joined #gnucash
06:49:57 *** Jimraehl1 has joined #gnucash
06:50:44 *** chris has joined #gnucash
06:51:10 *** Jimraehl1 has left #gnucash
06:55:24 *** chris has quit IRC
06:57:53 *** fabior has quit IRC
06:58:48 *** chris has joined #gnucash
07:21:52 *** chf has quit IRC
07:23:20 *** weasel has joined #gnucash
07:24:26 *** harshitaneja has quit IRC
07:24:43 *** chf has joined #gnucash
07:27:56 *** harshitaneja has joined #gnucash
07:32:53 *** harshitaneja has quit IRC
07:33:52 *** harshitaneja has joined #gnucash
07:37:03 *** weasel has quit IRC
07:44:11 *** jotrago has quit IRC
07:50:57 *** jiangfuqiao has quit IRC
07:58:51 *** oozer has joined #gnucash
08:01:29 *** chf has quit IRC
08:02:57 *** chf has joined #gnucash
08:09:26 *** weasel has joined #gnucash
08:21:13 *** jotrago has joined #gnucash
08:21:54 *** chf has quit IRC
08:27:32 *** fekepp has quit IRC
08:27:40 *** fekepp has joined #gnucash
08:30:45 *** fekepp has quit IRC
08:32:00 *** chf has joined #gnucash
08:34:33 *** ncv has quit IRC
08:36:09 *** fekepp has joined #gnucash
08:37:05 *** ncv has joined #gnucash
08:38:44 *** jiangfuqiao has joined #gnucash
08:38:44 *** ncv has joined #gnucash
08:39:51 *** NestorZ has joined #gnucash
08:41:42 <jiangfuqiao> hello
08:41:59 *** msekar has joined #gnucash
08:42:46 *** msekar has quit IRC
08:44:04 <chris> Hello!
08:46:37 *** Mechtilde has joined #gnucash
08:46:48 *** fabior has joined #gnucash
08:47:52 *** ncv has quit IRC
08:49:09 *** ncv has joined #gnucash
08:52:47 *** jiangfuqiao has quit IRC
08:53:46 *** chf has quit IRC
08:56:25 *** fekepp1 has joined #gnucash
08:56:41 *** fekepp has quit IRC
08:59:08 *** chf has joined #gnucash
08:59:28 *** fekepp1 has quit IRC
09:01:54 *** ncv has quit IRC
09:03:02 *** ncv has joined #gnucash
09:05:14 *** chf has quit IRC
09:13:38 *** ncv has quit IRC
09:14:02 *** chf has joined #gnucash
09:14:38 *** ncv has joined #gnucash
09:16:05 *** PaulW2U has quit IRC
09:20:30 *** mipesom has quit IRC
09:46:40 *** chf has quit IRC
09:50:25 *** Mechtilde has quit IRC
09:53:00 *** chf has joined #gnucash
09:53:56 *** benoitg has quit IRC
10:04:46 *** darshan has joined #gnucash
10:11:11 *** benoitg has joined #gnucash
10:12:48 *** NestorZ has quit IRC
10:13:51 *** chf has quit IRC
10:17:19 *** chf has joined #gnucash
10:20:50 *** darshan has quit IRC
10:26:15 *** chf has quit IRC
10:33:19 *** chf has joined #gnucash
11:16:39 *** KaiForce has joined #gnucash
11:17:50 *** bob_ has joined #gnucash
11:31:45 *** Mechtilde has joined #gnucash
12:11:49 *** bob_ has quit IRC
12:22:54 *** frakturfreak has joined #gnucash
12:43:44 *** frakturfreak has quit IRC
12:44:20 *** benoitg has quit IRC
13:37:03 *** ncv has quit IRC
13:40:46 *** jotrago1 has joined #gnucash
13:49:01 *** NestorZ has joined #gnucash
13:54:44 *** mipesom has joined #gnucash
13:55:21 *** KaiForce has quit IRC
14:01:56 *** frakturfreak has joined #gnucash
14:05:15 *** fabior has quit IRC
14:18:38 <gjanssens> jralls: there appear to be issues with recursive kvp frames still :(
14:19:11 <jralls> Sigh. Are you looking at a bug?
14:19:27 <gjanssens> I hit this by accident when trying to load a 2.6.18 created sqlite file in current unstable
14:19:58 <gjanssens> The "Num-for-action" feature was set in the old file and unstable refused to open it.
14:20:27 <gjanssens> Have been debugging the whole afternoon and for the features slots I understand what's going wrong
14:20:47 <gjanssens> My worry is similar issues could happen in all other slots
14:20:49 <jralls> Hmm, I thought that I'd fixed that.
14:21:09 <gjanssens> We have added two fixes related to slots in sql
14:21:29 <jralls> I meant the features slots not being recognized.
14:21:30 <gjanssens> One was because feature flags could have a leading "/"
14:21:46 <gjanssens> The other was for bayes, but that second fix broke features again
14:21:47 <jralls> Which is the one I was thinking of.
14:21:57 <jralls> Well, poop.
14:22:37 <gjanssens> The trouble is our notion of kvp frame name and path is no longer consistent across the code base
14:22:59 <gjanssens> For bayes we need to test on full path (minus leading slash)
14:23:08 <gjanssens> For features we should only test on key
14:23:57 <gjanssens> This is probably because our kvp frame hierarchy is only partly flattened (in particular for bayes data only)
14:24:39 <gjanssens> There probably are still holes in my understanding of all of this so it's not clear yet what the proper fix would be.
14:24:50 <gjanssens> Still working on that
14:25:31 <gjanssens> I also still have to study what the xml backend does in comparison.
14:25:45 <gjanssens> I'll try to get it ready for this WE as I think this is a blocker
14:26:29 <gjanssens> Do we have a sample file with bayes data in it ?
14:27:50 <jralls> test-dbi.xml has a couple of bayes entries. What it doesn't have and I guess needs is a features entry.
14:28:16 <gjanssens> That's quickly added by running the file once and setting the "num-for-action" feature
14:28:26 <gjanssens> I'll play with that one
14:38:25 <gjanssens> Hmm, it doesn't have bayes info. But I appear to have a few test books myself with such data.
14:40:23 *** jotrago has quit IRC
14:40:24 *** jotrago1 is now known as jotrago
14:40:44 <jralls> Yes it does, lines 132 and 136. I added them as part of e5561bd.
14:41:41 <jralls> Still, it would be nice to have a really extensive example file both for testing and generating screenshots.
14:42:23 *** benoitg has joined #gnucash
14:42:28 <gjanssens> Oh! I was looking at the maint version :/
14:43:07 <gjanssens> FWIW, the xml backend understands the features properly in both 2.6 and unstable
14:43:44 <gjanssens> So the issue is really in the sql backend.
14:44:18 <gjanssens> Which makes sense because the xml backend didn't store keys with separator like the sql backend does/did
14:45:34 *** benoitg has quit IRC
14:45:47 <jralls> Right. Backwards compatibility with my poor design is biting us.
14:46:24 <gjanssens> Eh, and the bayes slots in the test xml file are already flattened I suppose...
14:47:18 <jralls> Or maybe poor implementation of a poor design. The SQL backend wasn't supposed to store the paths either, it was supposed nest frames just like the XML backend. That results in poor load performance.
14:49:31 <gjanssens> Anyway, got enough samples now to go figure it out :)
14:49:52 <gjanssens> But it'll probably be for tomorrow. The jetlag is still active and I'm getting tired
14:50:40 <jralls> OK. We need to decide what to do about the release, but that can wait for tomorrow.
14:54:27 *** oozer has quit IRC
14:57:54 <jralls> gjanssens: About your comment on pr325, do you think that this problem is leaking into the representation of bayes KVP in memory?
15:05:47 *** benoitg has joined #gnucash
15:23:08 *** oozer has joined #gnucash
15:39:54 <gjanssens> jralls: I don't know for sure yet but I suspect so
15:40:12 <gjanssens> I'll know more tomorrow
15:40:16 <jralls> OK.
15:40:44 <gjanssens> jralls: I still have some time to discuss the release if you like
15:41:07 <gjanssens> What is still to be done ?
15:41:17 <gjanssens> Which bugs are blocking, which ones can wait ?
15:42:07 <jralls> Exactly. I didn't think any bugs were blocking, but we clearly need to sort this out first.
15:43:03 <jralls> Is there anything else you think should block release?
15:44:35 <gjanssens> There are a few crash reports still unhandled.
15:45:03 <jralls> Yes, but they're all Gtk bugs AFAICT.
15:45:05 <gjanssens> You have marked https://bugzilla.gnome.org/show_bug.cgi?id=779926 for 2.8. Do you plan to work on this ?
15:45:46 *** Mechtilde has quit IRC
15:46:57 <jralls> It's documentation, so I don't want to hold up the release for it.
15:49:28 <gjanssens> That's fine. I'm just going through bugs with target milestone 2.8 right now to see if there's something important left
15:53:24 *** NestorZ has quit IRC
15:54:48 <jralls> Separate subject, I found a trove of example files at https://www.gnucash.org/docs/examples/it. I think they're the ones cmarchi made for the documentation screenshots, which is good because I'm making new Gtk3 screenshots.
15:55:00 <jralls> What's weird is that they're not in git.
15:56:26 <gjanssens> That's weird indeed
16:07:02 <gjanssens> https://bugzilla.gnome.org/show_bug.cgi?id=794809 is a new crash reported today
16:08:03 <gjanssens> https://bugzilla.gnome.org/show_bug.cgi?id=794767 needs some investigation. Not begin able to enter business dates is serious
16:08:27 *** ArtGravity has joined #gnucash
16:10:08 <gjanssens> https://bugzilla.gnome.org/show_bug.cgi?id=793220 is this a blocking regression ?
16:15:08 <jralls> 794809 is mac-only and in Gtk. I couldn't reproduce 794767, can you? I also can't reproduce 793220, see my video on comment 8.
16:18:02 <gjanssens> I haven't tried yet. I'm just browsing now.
16:18:27 *** NestorZ has joined #gnucash
16:41:16 <gjanssens> :s:q
16:42:06 <gjanssens> Grmbl, my irc client insists on grabbing focus whenever I'm switching virtual screen :(
16:42:08 *** frakturfreak1 has joined #gnucash
16:42:53 *** frakturfreak has quit IRC
16:52:28 <gjanssens> https://bugzilla.gnome.org/show_bug.cgi?id=794791 is another one a arch in the transfer funds dialog
16:54:38 <gjanssens> Probably in the same category as bug 794730 and 794360...
16:56:33 <gjanssens> I think the only confirmed blocker at this point is the kvp issue I'm looking into
16:57:32 <gjanssens> The python bindings won't work due to the failing binreloc code as well, but I think we can release without those and mention it as a known issue in the release notes.
16:58:15 <gjanssens> And that's all I have for tonight.
16:58:29 <gjanssens> Is there anything else that's important to have ?
16:58:30 <jralls> 794791: Just tried to reproduce, it worked for me.
16:59:07 <gjanssens> Yes, I think it only fails on arch due to how they configure glib
16:59:25 <gjanssens> Did you get critical errors about invalid object types ?
16:59:31 <jralls> There are some documentation spruce ups. I've made some new screenshots but many of them will require setting up an example book.
16:59:53 <gjanssens> Ok
17:00:31 <gjanssens> Then it's time for me to call it a day.
17:00:52 <gjanssens> See you tomorrow!
17:01:37 <jralls> Good night!
17:02:11 <gjanssens> Oh, just a final note: I have played a bit more with gtk-3.0's settings.ini file
17:02:22 <jralls> Yeah, I did get criticals, "invalid uninstantiable type"
17:02:35 <gjanssens> On Fedora it is properly picked up and I can use it to change the default theme
17:02:52 *** harshitaneja has quit IRC
17:03:25 <gjanssens> This didn't work on Windows, but I haven't figured out yet whether gtk on windows is really looking for it. I haven't found the equivalent of strace yet to find which files a gnucash run tries to open/access
17:03:28 *** harshitaneja has joined #gnucash
17:03:56 <gjanssens> Definitely not critical, but so you know where I was with that
17:04:06 <jralls> What's wrong with strace? Mingw64 has it.
17:04:32 <jralls> Did you try setting XDG_CONFIG_HOME on windows?
17:04:39 <gjanssens> It does, but it doesn't show much, unless I used it improperly
17:04:56 <gjanssens> I haven't tried XDG_CONFIG_HOME yet
17:05:40 *** gjanssens is now known as gjanssens_afk
17:05:47 <jralls> The docs say that it looks in /etc, which it isn't going to find on Windows, then XDG_CONFIG_DIRS then XDG_CONFIG_HOME. No default to $HOME.
17:06:00 <jralls> Haven't looked at the code to see if that's true.
17:06:59 *** bob_ has joined #gnucash
17:07:48 <gjanssens_afk> Well, I put gtk-3.0/settings.ini in what is supposed to be the default for XDG_CONFIG_HOME on windows: AppData/Local
17:07:57 <gjanssens_afk> Also tried AppData/Roaming
17:08:12 <gjanssens_afk> But in either case there was no visual difference
17:09:02 <gjanssens_afk> So from there I figured to first determine where gtk actually looks, which lead me on a hunt for strace
17:09:13 <gjanssens_afk> without much actual results so far
17:09:43 <gjanssens_afk> Can be continued later, unless you like to dig into it
17:10:05 <jralls> There's always gdb, but you'd have to rebuild gtk to get symbols.
17:10:26 <jralls> Not hard, but a bit annoying.
17:10:35 <jralls> Anyway, off to bed with you! ;-)
17:31:12 *** NestorZ has quit IRC
18:23:46 *** mipesom has quit IRC
18:23:59 *** mipesom has joined #gnucash
18:24:53 *** benoitg has quit IRC
18:39:55 *** bob_ has quit IRC
18:47:45 *** harshitaneja has quit IRC
18:57:29 *** frakturfreak1 has quit IRC
19:54:11 *** chf has quit IRC
20:01:48 <fell> jralls: only if the distri provides no dbg packages.
20:02:45 *** gncbot sets mode: +o fell
20:04:31 <jralls> fell, got to go make dinner, but only what if the distro has no dbg pkg?
20:04:53 <fell> you have to rebuild gtk
20:05:43 <jralls> fell: Ah, yes, that was about Mingw-w64, which indeed provides no debug symbols.
20:07:25 <fell> Bon appetit!
20:37:20 *** chf has joined #gnucash
21:30:40 *** oozer has quit IRC
22:50:47 *** badger92 has quit IRC
23:04:32 <chris> I had a thought about the 'look' of the TR and friends -- we all expect the amounts in the TR to be blue and clickable... this is now deeply embedded into the Gnucash brand. The disadvantage is the html cell numbers become unusable when opened in Excel/Libreoffice because they point to useless gnc:register urls.
23:07:04 <chris> I can definitely modify this to move hyperlinks to a different column, and refine the display to provide html like <td sdval="2.00">$2.00</td> which means the cell numbers are directly available to spreadsheets. The SDVAL tag is not however an official standard.