2021-09-27 GnuCash IRC logs

00:36:08 *** Mechtilde has joined #gnucash
00:47:56 <jralls> warlord, commits aren't going through to www.gnucash.org: Ncat: Connection reset by peer.
00:49:11 <jralls> Two tries, one for Gnucash-Intel-4.7-2.dmg and a followup fake one to try and kick the connection.
00:56:26 <CDB-Man> Sorry for the delay folks. chris and jralls I've updated a new truth table to https://bugs.gnucash.org/show_bug.cgi?id=797796#c215 -- chris i have not had the time/opportunity to test your patch. will have to make time for it... next time, hopefully after you've also kicked the tires on the updated truth table
00:56:55 <CDB-Man> i forgot if i mentioned it to this group or not, but I've been promoted at work to manager! in our management consulting strategy practice
00:57:33 *** lyle has joined #gnucash
01:18:44 *** usamimikamen_sk has joined #gnucash
01:25:49 *** frakturfreak1 has quit IRC
01:26:53 *** ChanServ sets mode: +v usamimikamen_sk
01:40:19 *** frakturfreak1 has joined #gnucash
01:41:50 *** Mechtilde has quit IRC
01:44:05 *** usamimikamen_sk has quit IRC
01:45:43 *** Jeanl has joined #gnucash
01:46:38 *** jervin has quit IRC
01:47:23 *** Mechtilde has joined #gnucash
01:53:01 *** Mechtilde has quit IRC
01:58:09 *** jervin has joined #gnucash
02:03:12 *** jervin has quit IRC
02:20:36 *** joo has quit IRC
02:29:13 *** joo has joined #gnucash
02:29:13 *** ChanServ sets mode: +v joo
02:47:51 *** Jeanl has quit IRC
02:54:50 *** gjanssens has joined #gnucash
02:54:50 *** ChanServ sets mode: +o gjanssens
03:23:30 <chris> gjanssens: I found that the gsettings crash is not consistent - relaunching seems to be fine.
03:27:37 <gjanssens> chris: That depends on which gsettings issue you have. GnuCash will crash when it tries to use a GSettings schema that's not available.
03:29:11 <gjanssens> Migration more or less touches *all* schemas, so you're likely to hit a crash there but if it went far enough ahead to prevent a subsequent migration, true.
03:29:29 <gjanssens> If it's a schema that's used during normal gnucash operation, you will consistently get a crash.
03:30:15 <gjanssens> In this case it was a stupid copy/paste error on my behalf :(
04:23:45 *** User has joined #gnucash
05:27:04 *** sbluhm has joined #gnucash
05:27:04 *** Hamaryns has joined #gnucash
05:27:04 *** ChanServ sets mode: +v Hamaryns
06:07:12 *** Hamaryns has quit IRC
06:47:36 *** Jeanl has joined #gnucash
06:48:18 *** Jeanl has joined #gnucash
06:51:19 *** Jeanl has quit IRC
07:47:45 <warlord> jralls, is www up? That netcat is part of the port-knocker to get www to pull the site update. If that's not working, that will need linas to help debug
07:54:09 *** storyjesse has joined #gnucash
08:00:36 *** PhotoMan has quit IRC
08:00:51 *** PhotoMan has joined #gnucash
08:00:51 *** ChanServ sets mode: +v PhotoMan
08:03:56 <warlord> jralls, when I try it I get "already up to date" which implies that it DID pull.
08:04:11 <warlord> The connection reset by peer is normal.
08:04:25 <warlord> .. it closes as soon as the pull is complete.
08:22:10 *** storyjesse has quit IRC
09:29:21 *** sbluhm has quit IRC
09:32:58 *** jervin has joined #gnucash
09:42:45 *** jervin has quit IRC
09:42:55 *** jervin has joined #gnucash
09:48:22 *** jervin has quit IRC
09:49:42 *** jervin has joined #gnucash
10:10:52 *** jralls_laptop has joined #gnucash
10:18:34 *** jervin has quit IRC
10:30:57 *** hptina has joined #gnucash
10:39:36 *** ChanServ sets mode: +v hptina
10:42:35 *** KipITOne has joined #gnucash
10:42:36 *** ChanServ sets mode: +v KipITOne
10:42:58 *** jralls_laptop has quit IRC
10:43:02 *** jralls_laptop has joined #gnucash
10:43:23 *** jralls_laptop has joined #gnucash
10:43:28 *** bertbob has quit IRC
10:43:41 *** jralls_laptop has quit IRC
10:43:45 *** jralls_laptop has joined #gnucash
10:44:07 *** jralls_laptop has joined #gnucash
10:44:29 *** jralls_laptop has joined #gnucash
10:44:50 *** jralls_laptop has joined #gnucash
10:44:54 <hptina> hello everybody! :)
10:45:08 *** jralls_laptop has quit IRC
10:45:12 *** jralls_laptop has joined #gnucash
10:45:30 *** jralls_laptop has quit IRC
10:45:33 *** jralls_laptop has joined #gnucash
10:45:55 *** jralls_laptop has joined #gnucash
10:46:17 *** jralls_laptop has joined #gnucash
10:46:35 *** bertbob has joined #gnucash
10:46:35 *** ChanServ sets mode: +v bertbob
10:46:39 *** jralls_laptop has joined #gnucash
10:46:56 *** jralls_laptop has quit IRC
10:47:00 *** jralls_laptop has joined #gnucash
10:47:22 *** jralls_laptop has joined #gnucash
10:47:40 *** jralls_laptop has quit IRC
10:47:44 *** jralls_laptop has joined #gnucash
10:47:56 <hptina> I compiled gnucash 4.7 (KDE NEON == Ubuntu 20.04), but it crashes every time :(
10:48:02 *** jralls_laptop has quit IRC
10:49:13 <hptina> which information do you need in order to help me with that?
10:50:49 *** bertbob has quit IRC
10:52:28 *** bertbob has joined #gnucash
10:52:28 *** ChanServ sets mode: +v bertbob
11:00:32 <hptina> from the output of "make check": 34 - test-qof (Subprocess aborted)
11:00:32 <hptina> 57 - test-gnc-numeric (Failed)
11:02:25 *** jralls_laptop has joined #gnucash
11:03:08 <hptina> 'Use "--rerun-failed --output-on-failure" to re-run the failed cases verbosely.' <-- how to run that?
11:03:54 *** sbluhm has joined #gnucash
11:04:43 *** ChanServ sets mode: +o jralls_laptop
11:04:50 <jralls_laptop> hptina, Uninstall the old build if you have one and start again from an empty build directory.
11:06:02 <hptina> ok, I'll try that
11:06:02 <jralls_laptop> hptina the rerun-failed bit is for unit tests.
11:09:51 <hptina> "make -j4" is running ...
11:11:52 <hptina> btw, I'm running a custom kernel here
11:13:01 <jralls_laptop> The kernel shouldn't matter.
11:15:25 <hptina> ok :)
11:15:36 <jralls_laptop> I guess I should qualify that: The kernel shouldn't matter as long as GnuCash isn't crashing in a syscall.
11:16:18 <hptina> I don't hope so ...
11:16:33 <jralls_laptop> But so far the usual reason for 4.7 crashes is conflicts with the gsettings schema at build time. That's why I told you to uninstall and build in a clean build-dir.
11:18:16 <gjanssens> jralls_laptop: unfortunately there's also a typo in org.gnucash.GnuCash.deprecated.xml.gschema.in
11:18:29 <gjanssens> I have applied a patch to maint
11:18:47 <gjanssens> I only manifests during install, not during build.
11:18:56 <jralls_laptop> gjanssens Do we need to do a snap release?
11:19:19 <gjanssens> I suspect that has to do with the order in which the gschema files are processed by the schema file compiler
11:20:16 <gjanssens> During build, the aqbanking and ofx gschema files are explicitly copied before the main schemas. And hence the compiler already has the aqbanking new schema paths in cache
11:20:53 <gjanssens> However it looks as if sometimes during install the deprecated schema file is processed first and then it bails because of a missing schema reference.
11:21:24 <gjanssens> I don't know if it requires a snap release. I'm tempted to do one.
11:21:46 *** guak has joined #gnucash
11:21:53 <gjanssens> But otoh I haven't heard any Windows or MacOS user complain so far. Only self-builders
11:22:10 <gjanssens> Do you have detailed build logs for your macos release ?
11:22:57 <gjanssens> If so, does it contain this warning:
11:23:01 <gjanssens> Compiling gschema files in /home/janssege/Development/builds/gnucash/rel47/inst/share/glib-2.0/schemas
11:23:03 <gjanssens> Warning: undefined reference to <schema id='org.gnucash.GnuCash.dialogs.import.hbci.connection-dialog'/>
11:23:51 <gjanssens> That's the sign on linux of a bad build that will cause crashes at runtime.
11:24:09 *** sbluhm has quit IRC
11:24:54 <jralls_laptop> No, my computer crashed Saturday and wiped out all of the build records. But I haven't build on my laptop for a while so I can do a build now.
11:25:34 <gjanssens> warlord: re the www, it looks like jralls_laptop' last commit didn't even make it to code.gnucash.org. I have successfully updated the website this morning with the relevant version change and that is now up.
11:26:20 <gjanssens> That's probably why you didn't see any pull when you tried later today.
11:26:47 <warlord> Could be; I just ran the port-knock script by hand.
11:28:28 <gjanssens> jralls_laptop: I have now checked the build logs for the Windows release. It has the same warning, so it may crash as well.
11:30:21 <jralls_laptop> I think the problem with www is that I didn't save global_params when I committed the fixed dmg change on 210926-4.7.news, but that's the change I was looking for last night.
11:30:38 <gjanssens> Ok
11:31:33 <jralls_laptop> I'd think that if the 4.7 Windows or macOS build was crashing we'd have heard about it by now.
11:33:13 <jralls_laptop> On macOS building in a not-cleaned build-dir errors out with Error on line 3 char 1: <schema id='org.gnucash.dialogs.business'> already specified. --strict was specified; exiting.
11:33:59 <gjanssens> That's normal. You'll need to at least remove the schema files currently installed in your build directory.
11:35:25 <jralls_laptop> Yup, proceeding with that. The issue arises from installing into a $PREFIX with an existing install, right?
11:39:30 <gjanssens> No, I don't think so in this case.
11:40:07 <gjanssens> The issue is cmake for some reason doesn't update the schema files that are already in the *build* directory.
11:40:57 <gjanssens> However all schema files should be replaced as they all have org.gnucash replaced with org.gnucash.GnuCash in their schema nodes
11:41:19 <gjanssens> And a new file (the deprecated file) has the old schema nodes.
11:41:56 <gjanssens> Given the existing files aren't replaced, we end up with two files defining the same schema. The schema compiler doesn't like that.
11:42:17 <gjanssens> Possibly just explicitly rerunning cmake . in the build directory would have solved it as well.
11:42:57 <gjanssens> That all is why your most recent build failed.
11:43:44 <gjanssens> The issue encountered by others is that I had accidentally left one reference to an org.gnucash.GnuCash schema in the deprecated file (silly copy/paste error)
11:43:45 <jralls_laptop> I meant the issue with the crash, not the build failure.
11:44:31 <gjanssens> That also means the schema compiler complains about a reference to a missing schema, presumably because it processes that file first.
11:45:54 <gjanssens> I *suspect* it's not an issue on the Windows and Macos builds as they rerun the schema compiler while bundling and for some reason that one doesn't complain.
11:46:33 <gjanssens> I don't know why that is.
11:46:40 <jralls_laptop> Oh, I just did a pull so I've got your patch. That's not going to answer your question.
11:46:45 *** jervin has joined #gnucash
11:46:49 <gjanssens> Oops, no
11:47:36 *** jervin has quit IRC
11:48:36 *** o01eg has joined #gnucash
11:48:50 <gjanssens> Ok, just tested. The windows build doesn't fail and properly migrates
11:49:01 <jralls_laptop> Just restarted after checking out 4.7. ninja does re-run cmake after a git pull or checkout.
11:50:53 <gjanssens> Ok and does share/glib-2.0/schemas/org.gnucash.GnuCash.deprecated.gschema.xml contain a reference to 'org.gnucash.GnuCash' ?
11:51:00 <hptina> Thread 1 "gnucash" received signal SIGSEGV, Segmentation fault.
11:51:00 <hptina> 0x00007fffea6a95ec in ?? () from /usr/lib/x86_64-linux-gnu/gio/modules/libdconfsettings.so
11:51:11 <gjanssens> There should be a single one (which is the trigger of the crash)
11:52:16 <hptina> [Thread 0x7fffe91d5700 (LWP 305992) exited]
11:52:50 <hptina> gjanssens, a single what, please?
11:53:21 <gjanssens> hptina: that was a message to jralls_laptop :)
11:53:42 <hptina> ah, I see! ;)
11:53:46 <gjanssens> hptina: for you - you'll have to patch the tarball to get this fixed unfortunately
11:54:06 <gjanssens> You can find the patch at the end of this mailinglist message: https://lists.gnucash.org/pipermail/gnucash-user/2021-September/097769.html
11:54:55 <gjanssens> It's already applied to our git repository, but after the release commit.
11:55:33 <hptina> gjanssens, thank you very much; I'll try that patch! :)
11:56:48 <gjanssens> jralls_laptop: re trigger of the crash, it's really causing the schema compilation to abort early or to skip certain files for processing. And that evenutally leads to the crash as gnucash won't find all needed schemas at runtime.
11:56:53 *** chris has quit IRC
11:57:21 *** sergiomiguelrp has joined #gnucash
12:00:38 *** jervin has joined #gnucash
12:03:49 <jralls_laptop> gjanssens I do get an error in install but it's a little different: https://gist.github.com/jralls/15bb762f0717c22a04fdc61900022f88
12:08:40 *** ArtGravity has joined #gnucash
12:08:41 *** ChanServ sets mode: +v ArtGravity
12:11:15 *** NoobAlice has joined #gnucash
12:13:32 <gjanssens> jralls_laptop: that one means there is still a conflict between some schema files on your system.
12:14:36 <gjanssens> The error indicates the old schema file is still in your install directory, not the updated one.
12:14:44 <jralls_laptop> Well, yes, I deliberately left the old schema files in the installation dir, I thought that that was what prompted the problem.
12:15:06 <jralls_laptop> I'll remove them and install again.
12:15:40 <gjanssens> Oh, sorry, *this* indeed means you have both the old and the new file in there
12:16:00 <gjanssens> I already forgot I renamed the schema file in addition to updating their content
12:16:34 <gjanssens> So you can have an org.gnucash.dialogs.business.gschema.xml and an org.gnucash.GnuCash.gschema.xml
12:16:52 <gjanssens> It's indeed a matter of having an unclean prefix
12:18:41 <jralls_laptop> And that produces the "Warning: undefined reference to <schema id='org.gnucash.GnuCash.dialogs.import.hbci.connection-dialog'/>" that you expected.
12:19:20 <gjanssens> Right
12:19:47 <jralls_laptop> "unclean prefix" sounds like a religious problem. ;-)
12:19:58 <gjanssens> LOL
12:21:05 *** Simon has quit IRC
12:21:17 *** Simon has joined #gnucash
12:21:17 *** ChanServ sets mode: +v Simon
12:39:17 <jralls_laptop> BTW it looks like the schema files are generated at configure instead of being a build target. Is that intentional, accidental, or just unavoidable?
12:41:47 *** o01eg has quit IRC
12:43:41 *** Mechtilde has joined #gnucash
12:44:07 <gjanssens> That's intentional and should work fine.
12:45:13 <gjanssens> cmake's configure_file command adds the necessary bits in the generaded build system to update the generated schema files whenever a source file changes.
12:45:20 <hptina> hmm, gnucash crashed again, after applying the patch and freshly building :((
12:45:34 <gjanssens> That's the rerun of cmake you sometimes see when starting ninja
12:45:42 <gjanssens> Huh crash with what error message ?
12:46:08 <hptina> sigsegv
12:46:36 <jralls_laptop> Where?
12:46:43 <gjanssens> hptina: then it's probably not the gschema error. Those crash with sigabrt
12:47:27 <hptina> well, I'll run it again after "make install"
12:48:36 <gjanssens> Is your install directory empty as well ? Or more precisely did you remove all files starting with org.gnucash.* from <inst_prefix>/share/glib2.0/schemas ?
12:49:39 <hptina> I restarted with an empty build directory
12:50:13 <gjanssens> And it crashed when running from that build directory ?
12:53:37 <hptina> gdb says: Thread 1 "gnucash" received signal SIGTRAP, Trace/breakpoint trap.
12:53:38 <hptina> 0x00007ffff7b50295 in ?? () from /lib/x86_64-linux-gnu/libglib-2.0.so.0
12:54:20 <hptina> the same error as without that recent patch
12:57:09 <hptina> this output comes from gnucash after "make install"
12:58:03 <gjanssens> hptina: ok, SIGTRAP is indeed how missing schema errors cause gnucash to abort.
12:58:24 <gjanssens> What's your install prefix ?
12:58:37 *** jervin has quit IRC
12:58:41 <hptina> /usr/local
12:59:20 <gjanssens> Ok, what files do you have in /usr/local/share/glib2.0/schemas that start with org.gnucash. ?
13:00:56 <gjanssens> Do you have file pairs of which one is likeorg.gnucash.<something>.gschema.xml and the other like org.gnucash.GnuCash.<something>.gschema.xml ?
13:01:07 <hptina> CMakeCache.txt:CMAKE_CXX_FLAGS:STRING=-march=native -pipe
13:01:07 <hptina> CMakeCache.txt:CMAKE_C_FLAGS:STRING=-march=native -pipe <-- should I try again without those flags?
13:01:08 <gjanssens> (So just an added GnuCash. in the f ilename)
13:01:40 <hptina> one moment, please
13:05:53 <hptina> org.gnucash.GnuCash.window.pages.gschema.xml
13:05:53 <hptina> org.gnucash.GnuCash.window.pages.account.tree.gschema.xml
13:05:53 <hptina> org.gnucash.GnuCash.warnings.gschema.xml
13:05:53 <hptina> org.gnucash.GnuCash.history.gschema.xml
13:05:53 <hptina> org.gnucash.GnuCash.gschema.xml
13:05:56 <hptina> org.gnucash.GnuCash.general.finance-quote.gschema.xml
13:05:58 <hptina> org.gnucash.GnuCash.dialogs.totd.gschema.xml
13:06:00 <hptina> org.gnucash.GnuCash.dialogs.sxs.gschema.xml
13:06:02 <hptina> org.gnucash.GnuCash.dialogs.reconcile.gschema.xml
13:06:04 <hptina> org.gnucash.GnuCash.dialogs.import.qif.gschema.xml
13:06:06 <hptina> org.gnucash.GnuCash.dialogs.import.ofx.gschema.xml
13:06:08 <hptina> org.gnucash.GnuCash.dialogs.import.hbci.gschema.xml
13:06:10 <hptina> org.gnucash.GnuCash.dialogs.import.generic.gschema.xml
13:06:12 <hptina> org.gnucash.GnuCash.dialogs.import.csv.gschema.xml
13:06:14 <hptina> org.gnucash.GnuCash.dialogs.gschema.xml
13:06:16 <hptina> org.gnucash.GnuCash.dialogs.flicker.gschema.xml
13:06:18 <hptina> org.gnucash.GnuCash.dialogs.export.csv.gschema.xml
13:06:20 <hptina> org.gnucash.GnuCash.dialogs.commodities.gschema.xml
13:06:22 <hptina> org.gnucash.GnuCash.dialogs.checkprinting.gschema.xml
13:06:26 <hptina> org.gnucash.GnuCash.dialogs.business.gschema.xml
13:06:28 <hptina> org.gnucash.GnuCash.deprecated.gschema.xml
13:06:30 <hptina> sorry for flooding!
13:07:50 <gjanssens> And none that are only org.gnucash.<something>.gschema.xml ?
13:08:21 <gjanssens> like org.gnucash.gschema.xml
13:08:56 <hptina> no, none
13:09:21 <gjanssens> Ok, back to your SIGTRAP. Did it come with another message right before by GIO?
13:09:41 <gjanssens> Something like '* 18:56:53 OTHER <GLib-GIO > Settings schema 'bla' is not installed' ?
13:10:02 <hptina> I didn't notice something like that
13:10:56 <gjanssens> Oh, probably not. I have set my log to go to terminal. But you may have to log in /tmp/gnucash.trace
13:11:16 <gjanssens> I mean *look* in /tmp/gnucash.trace
13:11:24 <hptina> I'll take a look ...
13:12:21 <hptina> * 18:52:45 WARN <gnc.gui> [gnc_load_app_icons()] No icon named 'gnucash-icon' found. Some gui elements may be missing their icons
13:12:21 <hptina> * 18:52:46 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_bool()] Unable to set value for key show-at-startup in schema dialogs.totd
13:12:21 <hptina> * 18:52:46 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_int()] Unable to set value for key current-tip in schema dialogs.totd
13:12:21 <hptina> * 18:52:46 OTHER <GLib-GIO> Settings schema 'org.gnucash.GnuCash.dialogs.import.hbci' is not installed
13:13:13 <gjanssens> Ok, so clearly you are experiencing the GSettings bug.
13:13:36 <gjanssens> Did you also try to run from your build directory after applying the patch ?
13:13:49 <hptina> hmm, I haven't installed libchipcard yet - does that maybe matter?
13:14:11 <hptina> yes, with sigsegv
13:14:12 <gjanssens> No, not at all
13:14:23 <gjanssens> So let's go backwards.
13:14:54 <gjanssens> Go to <your-build-dir>/share/glib2.0/schemas
13:15:04 <hptina> ok
13:15:41 <gjanssens> Are there any org.gnucash.<something>.gshema.xml files in there ? So without the 'GnuCash.' bit in between the name.
13:16:36 <hptina> no, only "gschemas.compiled"
13:17:13 <gjanssens> Ok. But the files named org.gnucash.GnuCash.<something>.gschema.xml are there ?
13:17:22 <hptina> yes
13:18:06 <hptina> 21 files
13:19:06 <gjanssens> Good. What does this command (in that directory) return: grep 'org.gnucash.GnuCash' org.gnucash.GnuCash.deprecated.gschema.xml
13:20:03 <hptina> no output
13:21:52 <gjanssens> Ok. Which file is more recent ? gschemas.compiled or org.gnucash.GnuCash.deprecated.gschema.xml ?
13:25:00 <hptina> gschemas.compiled is about 4 minutes younger
13:30:26 <gjanssens> Ok, can you restart gnucash from your build directory to get more details on the crash ?
13:32:15 <hptina> * 19:31:03 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_bool()] Unable to set value for key show-at-startup in schema dialogs.totd
13:32:15 <hptina> * 19:31:03 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_int()] Unable to set value for key current-tip in schema dialogs.totd
13:32:15 <hptina> * 19:31:03 WARN <gnc.core-utils> Could not kill child process: Kein passender Prozess gefunden
13:32:35 *** kcin has joined #gnucash
13:33:39 <gjanssens> And did it crash ?
13:34:29 <hptina> yes
13:34:34 <gjanssens> Because this is a different error
13:34:39 <hptina> always :(
13:34:41 <gjanssens> Where did it crash ?
13:35:36 <hptina> it crashes shortly after starting the program
13:37:12 <hptina> I only get the "Tip of the day" on the screen, very shortly (< 1sec:)
13:39:20 <gjanssens> Can you provide a backtrace ?
13:39:28 <gjanssens> (BRB, called for supper)
13:40:44 <hptina> hmm, how to generate such a backtrace?
13:41:29 <hptina> with strace?
13:48:18 <jralls_laptop> With gdb.
13:49:01 <jralls_laptop> See https://wiki.gnucash.org/wiki/Stack_Trace.
13:49:46 *** jralls_laptop has quit IRC
13:55:35 <hptina> wait please ...
14:10:29 <hptina> here is the backtrace: https://gist.github.com/hptina-bm/89a4830a2e68113c7fee33833ebe17f8
14:11:48 <jralls> Hmm, still in aqbanking. Are you sure you applied gjanssens's patch correctly?
14:12:50 <hptina> patch -p1 < attachment-0001.bin
14:14:09 <gjanssens> And what is the output in gnucash.trace for this crash ? It looks like this is the error when run from /usr/local, not from your build directory.
14:14:10 <hptina> I didn't get any error message from that
14:14:49 <gjanssens> Considering the previous grep command didn't return anything I believe your patch was applied correctly.
14:14:56 <hptina> no, I executed in from within build
14:15:17 <gjanssens> Ok, so what's in gnucash.trace ?
14:17:14 <hptina> grep import.hbci.connection-dialog ../gnucash/gschemas/org.gnucash.GnuCash.deprecated.gschema.xml.in
14:17:14 <hptina> <child name="connection-dialog" schema="org.gnucash.dialogs.import.hbci.connection-dialog"/>
14:17:14 <hptina> <schema id="org.gnucash.dialogs.import.hbci.connection-dialog" path="/org/gnucash/dialogs/import/hbci/connection-dialog/">
14:17:38 <hptina> that's the patched file
14:18:22 <gjanssens> I believe you, but that's the info I need right now :)
14:18:24 <hptina> * 19:51:48 WARN <gnc.gui> [gnc_load_app_icons()] No icon named 'gnucash-icon' found. Some gui elements may be missing their icons
14:18:24 <hptina> * 19:51:49 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_bool()] Unable to set value for key show-at-startup in schema dialogs.totd
14:18:24 <hptina> * 19:51:49 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_int()] Unable to set value for key current-tip in schema dialogs.totd
14:18:25 <hptina> * 19:51:49 OTHER <GLib-GIO> Settings schema 'org.gnucash.GnuCash.dialogs.import.hbci' is not installed
14:18:39 <gjanssens> Ok
14:18:45 <gjanssens> That is
14:19:04 <gjanssens> Do you have another gnucash installed in /usr/bin ?
14:19:12 <hptina> no
14:19:52 <hptina> whereis gnucash
14:19:52 <hptina> gnucash: /usr/local/bin/gnucash /usr/local/etc/gnucash /usr/local/lib/gnucash
14:20:13 <jralls> Doesn't matter where it's installed, if it's installed on the paths including ld.so.conf ones then those get loaded instead of the build dir ones.
14:20:51 <gjanssens> That's an important clue
14:21:02 <jralls> There are environment variables that will override that for loadable modules and guile scripts, but they don't affect shared libraries.
14:21:31 <jralls> I don't think they affect gsettings schemas either.
14:22:02 <gjanssens> gsettings schemas are search for in all directories in XDG_DATA_DIRS
14:23:13 <jralls> Which wouldn't normally include build dir, so we must have a hack in place.
14:23:31 <jralls> Does it look in build dir first or last?
14:23:45 <gjanssens> We so update XDG_DATA_DIRS if the installation path is not a default path
14:24:30 <gjanssens> So yes our environment file will have an XDG_DATA_DIRS entry for the build dir output files
14:25:37 <gjanssens> It's found in <builddir>/etc/gnucash/environment
14:26:01 <gjanssens> It probably won't be there in case of installing in /usr/local
14:26:31 <gjanssens> IOW /usr/local/etc/gnucash/environment should not have an XDG_DATA_DIRS entry.
14:26:49 <gjanssens> I doubt it matters though.
14:27:41 <hptina> well, I found a thread in the WWW where somebody had a similar issue with XDG_DATA_DIRS, but the solution didn't fit to my issue
14:28:41 <jralls> Anyway, run ninja uninstall and then try running from the build dir again.
14:28:45 *** sbluhm has joined #gnucash
14:28:56 <gjanssens> hptina: what's your build directory path ?
14:29:14 <hptina> hmm, I don't build with ninja
14:29:44 <hptina> ~/src/gnucash/build
14:29:50 <jralls> OK, make uninstall then.
14:30:00 <hptina> ok
14:31:29 <hptina> done
14:32:13 <jralls> And?
14:32:31 <hptina> and now?
14:33:11 <gjanssens> Well, did you rerun gnucash from the build dir and did it crash ?
14:33:28 <hptina> wait, please ...
14:34:00 <hptina> crash! :((
14:34:31 <hptina> * 20:33:41 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_bool()] Unable to set value for key show-at-startup in schema dialogs.totd
14:34:31 <hptina> * 20:33:41 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_int()] Unable to set value for key current-tip in schema dialogs.totd
14:36:01 <gjanssens> Ok. Can you post the file ~/src/gnucash/build/share/glib2.0/gschemas.compiled somewhere I can download it ?
14:38:45 <gjanssens> Also I suspect you must have another problem because the two errors you get before the crash suggest GSettings can find some schemas but can't write to them
14:40:35 <hptina> backtrace: https://gist.github.com/hptina-bm/f33b907674db3b5022af2d4665d1188c
14:41:00 <hptina> one moment, please ...
14:43:57 <jralls> Uh-oh. You have a problem in libdconfsettings.
14:45:44 <jralls> Can you install the dbg packages for glib-2.0 and dconf and make another stack trace?
14:47:52 <hptina> https://www.file-upload.net/download-14703045/gschemas.compiled.html
14:51:58 <hptina> I can't find such a dbg package :(
14:58:32 <hptina> hmm, do I need the development package for libdconf?
15:03:53 <hptina> gjanssens: https://www.file-upload.net/download-14703045/gschemas.compiled.html
15:06:47 <jralls> Devel packages provide the headers you need to compile programs using the library. Debug (or dbg) packages provide the library's internal symbols and make for more detailed stack traces. That's what we need to see why libdconfsettings is segfaulting.
15:07:53 <jralls> What distro are you using?
15:08:52 <hptina> jralls, I know; you mentioned dconf so I thought I may need libdconf-dev
15:09:43 <hptina> I have KDE Neon here (= Ubuntu Focal 20.04)
15:09:55 <jralls> No, you need libdconf-dbg or whatever your distro calls the debug symbols package.
15:10:48 <hptina> I understand, but it isn't available
15:12:17 <hptina> hmm, Gnome is the hell on earth
15:13:21 <jralls> I dunno that it's any worse than any of the other cross-platform toolkits, Qt included.
15:14:12 <warlord> It's not gnome per se, hptina -- it's just the way software is packaged.
15:14:59 <gjanssens> hptina: you need to enable extra softwar repositories to be able to install debug packages
15:15:10 <gjanssens> Here's one link to explain how to do that: https://askubuntu.com/questions/487222/how-to-install-debug-symbols-for-installed-packages#487567
15:15:28 <hptina> hmm, isn't dconf a subproject of gnome?
15:18:44 <gjanssens> Not necessarily, it's a cross-desktop configuration storage layer built on top of dbus
15:21:22 <gjanssens> Anyway, the gsettings.compiled file indicates you do have the schema the earlier GIO error claimed you were missing
15:21:27 <gjanssens> org.gnucash.GnuCash.dialogs.import.hbci is in there.
15:22:08 <gjanssens> That's also confirmed by the fact you now have a different crash and the error messages.
15:24:09 <jralls> The package in question isn't dconf it's dconf-gsettings-backend, though it may turn out that you need the dconf symbols too; the debugger might not be able to see past libdconfsettings because its symbols are stripped.
15:24:18 *** fell has joined #gnucash
15:24:18 *** ChanServ sets mode: +o fell
15:28:08 <hptina> hmm, I tried to set up my repos in order to install those dbg packages, but synaptic throws out an ugly error :(
15:28:58 <hptina> brb
15:31:14 <jralls> Did you quit synaptic while you modified /etc/apt/sources.list?
15:32:10 <gjanssens> jralls: meanwhile how serious do you estimate this issue ? Should we do a snap release for it?
15:32:44 <gjanssens> Windows and Macos are unaffected. All selfbuiliders probably will be affected
15:33:31 <gjanssens> I have no idea how it will act when packaged in rpms or debs or whatever is out there.
15:34:18 <gjanssens> All these packages will recompile the schema files during installation, but I can't tell whether that will trip over the bad reference or not.
15:34:33 <jralls> I'm inclined to think that we don't. IIUC the packagers all build from absolute scratch, as does Flathub (which I've gotten to build succsssfully BTW). So as long as it's just clear out the schemas from prefix and do a clean build I don't think we need to do a re-release.
15:35:24 <jralls> But if hptina's crash indicates that there's something more then I might change my ming.
15:35:28 <jralls> err, mind.
15:36:21 <jralls> FWIW I tested aqbanking setup from the 4.7 dmg this morning, no problem.
15:37:55 <gjanssens> There's two issues
15:38:16 <gjanssens> 1. you need to build in a clean environment to not pick up the old schema files
15:38:24 <gjanssens> That everybody can do
15:38:35 <jralls> rpms and debs will remove the old everything and replace it with the new everything, so once they're built it shouldn't impact users.
15:39:17 <hptina> »http://ddebs.ubuntu.com focal-security Release« release file is missing ...
15:39:18 <gjanssens> 2. the bad reference in org.gnucash.GnuCash.gschema.xml makes that during installation (on linux) gschemas.compiled is incomplete, causing a crash at startup
15:40:04 <gjanssens> It's this part I don't know how it will behave when installing the rpm or deb
15:51:08 *** Mechtilde has quit IRC
15:52:50 <jralls> Are you sure about #2? I'd think that would apply to flatpak as well and the gschemas.compiled in my flatpak build has hbci.connection-dialog.
15:52:56 <gjanssens> jralls: I think I have found what makes the difference between self-building and packaged builds
15:53:32 <gjanssens> Our cmake command runs glib-compile-schemas with the --strict option
15:53:43 <gjanssens> Our packaging scripts don't
15:54:31 <jralls> But flatpak...
15:54:34 <gjanssens> If I try to run a cleanly built gnucash directly from where it gets installed with ninja-build install (and without the patch applied) it crashes
15:54:58 <gjanssens> If I manually run glib-compile-schemas without --strict on the installed schemas, gnucash works
15:58:24 <gjanssens> I can only assume flatpak itself run glib-compile-schemas as part of its voodoo.
15:58:59 <gjanssens> That would make sense as glib-compile-schemas must be run on the whole of schema files installed by all components in the package
15:59:03 <gjanssens> Not only the gnucash ones.
15:59:21 <gjanssens> That's why we also run it as part of our bundle script
15:59:35 <gjanssens> (on Windows that is, and I presume on Macos as well)
16:00:11 <gjanssens> I added that --strict flag myself recently for exactly the opposite reason, but that doesn't seem to do what we want:
16:01:16 <gjanssens> I wanted --strict to stop the build in case duplicate schema were found, but it clearly doesn't. It just stops further processing of gschema processing.
16:05:44 <jralls> I get the same crash on macOS when run from the install dir without bundling. How do I run an unpublished flathub build?
16:07:25 <jralls> Hmm, and hbci.connection-dialog is in gschemas.compiled, so that's not a good indication of the problem.
16:12:20 <jralls> Oh, of course: It's still finding it in the old schema and the crash comes when it tries to copy it to the new path because the path isn't there.
16:15:22 <gjanssens> Yes, that's what it's about.
16:16:19 <gjanssens> So you have to check whether org.gnucash.GnuCash.dialogs.import.hbci is present in gschemas.compiled
16:16:40 <gjanssens> I used the "strings" command to just extract the strings from the binary file.
16:17:07 <gjanssens> In the build dir this string is in the file, but in the install dir that string is missing
16:17:26 <gjanssens> Together with a lot of other strings
16:18:39 *** field^Zzz3 has joined #gnucash
16:18:48 <jralls> Ah, it's the together with a lot of other strings because --strict bails out.
16:18:56 <gjanssens> Yes
16:19:45 <gjanssens> But here's where it gets weird: if you rerun glib-compile-schemas on that directory a second time it works -- regardless of whether you set the --strict paremeter or not.
16:20:21 <jralls> So how to test the Flathub build? It's committed but not yet published. And for that matter how do I keep it from publishing?
16:21:29 <jralls> Is there a flathub control panel somewhere?
16:21:55 <gjanssens> I only know of the page where you can follow the builds
16:22:29 <gjanssens> But I thought there is a way to install an unpublished flatpak for evaluation
16:24:12 <jralls> I'm sure there is, but there's very little documentation.
16:30:16 *** sbluhm has quit IRC
16:30:37 <gjanssens> jralls: It's at the end of the build log on flathub's buildbot:
16:30:42 <gjanssens> flatpak install https://dl.flathub.org/build-repo/59187/org.gnucash.GnuCash.flatpakref
16:31:09 <gjanssens> (found on this page: https://flathub.org/builds/#/builders/32/builds/61353)
16:31:55 <jralls> Thanks, installing.
16:39:34 <jralls> And crash.
16:39:43 <jralls> So we need to do a snap release.
16:40:48 <jralls> But we need to be sure that it will work.
16:41:58 <gjanssens> You can wait for our own nightly to pick up my patch
16:42:58 <gjanssens> Anyway, it's my time to leave for the night. Hopefully you have enough information to continue to work on this ?
16:43:45 <jralls> Yes, I think so. I'm concerned about hbtina's segfault, but I think we both know about the same about that.
16:45:22 <gjanssens> hptina's segfault is not the same, that's for sure. But we don't know if it's another side effect of this.
16:46:36 <gjanssens> I doubt it as the logs first show two write errors. If you follow our code those errors are only printed in the log if the schema and key are already validated and only the writing fails.
16:47:39 *** User has quit IRC
16:47:41 <gjanssens> But I may be jumping to conclusions here. The write attempts come from the tip of the day dialog, the crash happens when gnucash attempts to reset the temporary warnings.
16:47:49 <jralls> It doesn't matter, we need to understand it. If it's something we did then we need to fix it before doing a snap release.
16:47:56 <gjanssens> True
16:48:53 <jralls> It's crashing in g_settings_reset, apparently on the dialog.totd key.
16:49:49 <gjanssens> I don't think it's on the dialog.totd key
16:50:02 <gjanssens> It's later in the code, when gnucash tries to reset temporary warnings
16:50:24 <gjanssens> That's the only spot where we call gnc_prefs_reset_group
16:50:52 <gjanssens> But the interesting bit is even then the schema and key are first verified to be valid
16:51:21 <gjanssens> So it rather looks as if there's a write problem in dconf than there is a problem in our schema installation
16:51:53 <gjanssens> So g_settings_set_bool and g_settings_set_int fail with an error, g_settings_reset segfaults
16:52:19 <gjanssens> But both schema and key are checked to be valid before these functions are called.
16:53:34 <gjanssens> hptina: can you run dconf-editor on your system and try to make changes to the gnucash preferences that way ?
16:54:27 <gjanssens> For that to work you will probably have to set environment variable XDG_DATA_DIRS to <yourbuilddir>/share:$XDG_DATA_DIRS
16:54:40 <gjanssens> Otherwise our gnucash schema won't be picked up.
16:54:55 <gjanssens> That's all extra I can offer tonight.
16:54:58 <gjanssens> Good luck !
16:55:22 <jralls> I don't see your patch in maint. Are you sure you pushed it?
16:55:34 <gjanssens> Oh let me check
16:57:43 <gjanssens> Yes: https://github.com/Gnucash/gnucash/commit/b7001b52cf2842f056b7ad919791734462f2b1a4
16:58:26 <gjanssens> The macos test failed due to an expired certificate, but that seems unrelated...
16:59:08 <jralls> https://sourceforge.net/p/forge/site-support/22740/
16:59:15 <gjanssens> Do you intend to repackage our flatpak with only that patch added ?
16:59:32 <gjanssens> You could to avoid having to to a complete snap install
16:59:39 <gjanssens> to *do*
17:00:31 <jralls> No, if the flatpak fails I have to assume that rpms, debs, etc. will too. I'll release 4.8.
17:01:04 <gjanssens> Ok, sorry for the trouble :(
17:01:31 <gjanssens> I really am leaving this time :)
17:01:35 <gjanssens> Bye!
17:01:41 <jralls> Goodnight!
17:01:43 *** gjanssens has quit IRC
17:08:32 *** kcin has quit IRC
17:26:22 *** o01eg has joined #gnucash
17:35:26 <jralls> hptina since you can't seem to get the dbg packages can you do a little printf-debugging?
17:36:33 <hptina> jralls, how to do that?
17:37:53 <jralls> Open libgnucash/app-utils/gnc-gsettings.cpp in your favorite editor and find the function gnc_settings_reset at line 551.
17:37:57 <hptina> sorry, I was afk for a while ...
17:38:20 <hptina> I installed dconf-editor right now
17:39:07 <jralls> OK, but first I want to find out what key is tripping the crash.
17:39:37 <jralls> Then you can try changing it in dconf-editor and see if that crashes too.
17:40:49 <hptina> what and how to edit in line 551?
17:41:04 <jralls> Don't you have a text editor?
17:41:29 <hptina> already open (vim) ;)
17:42:11 <jralls> OK. So in that function there's if (gnc_gsettings_is_valid_key...). After that line insert a {
17:42:37 <jralls> Then insert PWARN("Resetting key %s", key);
17:43:02 <jralls> Then skip a line and put a } just above the else
17:43:13 <jralls> save, recompile, and crash.
17:43:52 <jralls> The last line of gnucash.trace will name the key that crashed. Maybe crash it 2 more times to make sure its always the same key.
17:44:33 <hptina> if (gnc_gsettings_is_valid_key (settings_ptr, key))
17:44:34 <hptina> 558 {
17:44:34 <hptina> 559 PWARN("Resetting key %s", key);
17:44:34 <hptina> 560 } <-- is that correct?
17:44:44 *** o01eg has quit IRC
17:45:15 <jralls> no, g_settings_reset (settings_ptr, key); needs to go before the closing }
17:45:47 <jralls> so
17:45:57 <jralls> 558 {
17:46:06 <jralls> 559 PWARN("Resetting key %s", key);
17:46:24 <jralls> 560 g_settings_reset (settings_ptr, key);
17:46:55 <jralls> 561 }
17:47:05 <jralls> 562 else
17:48:13 <hptina> ok, got it :)
17:49:17 <hptina> "make -j4" is running ...
17:52:27 <hptina> hmm, how to work with dconf-editor?
17:56:10 <jralls> That I don't know. What was the key?
17:56:46 *** field^Zzz3 has quit IRC
17:57:48 <hptina> idk :(
17:58:39 <jralls> Did you crash gnucash again and look at the bottom of gnucash.trace?
17:59:46 <hptina> it's still compiling
18:00:11 <jralls> Oh, OK. So "not yet" rather than "idk".
18:01:21 <hptina> * 00:00:13 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_bool()] Unable to set value for key show-at-startup in schema dialogs.totd
18:01:21 <hptina> * 00:00:13 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_int()] Unable to set value for key current-tip in schema dialogs.totd
18:01:21 <hptina> * 00:00:13 WARN <gnc.app-utils.gsettings> [gnc_gsettings_reset()] Resetting key reg-is-acct-pay-rec
18:02:57 <jralls> Interesting. The first one.
18:04:00 <jralls> So fire up dconf-editor. Click org in the window, then gnucash then GnuCash.
18:05:05 *** waitman has quit IRC
18:05:12 *** hptina has quit IRC
18:05:12 *** warlord has quit IRC
18:05:12 *** CDB-Man has quit IRC
18:05:12 *** ericdm has quit IRC
18:05:12 *** tg_bot1 has quit IRC
18:05:12 *** raeburn has quit IRC
18:05:12 *** fiddlerwoaroof has quit IRC
18:05:12 *** immae1 has quit IRC
18:05:12 *** prg has quit IRC
18:05:12 *** NoobAlice has quit IRC
18:05:12 *** bertbob has quit IRC
18:05:12 *** PhotoMan has quit IRC
18:05:12 *** Pegasus_RPG has quit IRC
18:05:12 *** chf1 has quit IRC
18:05:12 *** PowaBanga has quit IRC
18:05:12 *** TownsendHardware has quit IRC
18:05:12 *** g5pw has quit IRC
18:05:12 *** ldir has quit IRC
18:05:12 *** ekleog_ has quit IRC
18:05:12 *** crossan007 has quit IRC
18:05:13 *** waitman has joined #gnucash
18:05:27 *** NoobAlice has joined #gnucash
18:05:27 *** bertbob has joined #gnucash
18:05:27 *** hptina has joined #gnucash
18:05:27 *** PhotoMan has joined #gnucash
18:05:27 *** Pegasus_RPG has joined #gnucash
18:05:27 *** warlord has joined #gnucash
18:05:27 *** CDB-Man has joined #gnucash
18:05:27 *** ericdm has joined #gnucash
18:05:27 *** chf1 has joined #gnucash
18:05:27 *** tg_bot1 has joined #gnucash
18:05:27 *** PowaBanga has joined #gnucash
18:05:27 *** TownsendHardware has joined #gnucash
18:05:27 *** raeburn has joined #gnucash
18:05:27 *** g5pw has joined #gnucash
18:05:27 *** ldir has joined #gnucash
18:05:27 *** fiddlerwoaroof has joined #gnucash
18:05:27 *** immae1 has joined #gnucash
18:05:27 *** ekleog_ has joined #gnucash
18:05:27 *** prg has joined #gnucash
18:05:27 *** crossan007 has joined #gnucash
18:05:27 *** irc.eagle.y.se sets mode: +vvvqovvv bertbob hptina PhotoMan warlord warlord CDB-Man g5pw ldir
18:06:37 <hptina> jralls, did you get my input?
18:09:01 <jralls> I saw what you pasted from gnucash.trace and noted that it was crashing on the first key.
18:09:14 <hptina> ok
18:10:35 <jralls> Did you crash it more than once to make sure it always crashes on the same key?
18:11:05 <hptina> twice - both with the same output
18:12:37 <jralls> OK. Try installing GnuCash and running it from /local/bin.
18:13:15 <hptina> ok
18:14:36 <hptina> */usr/local/bin
18:15:35 <jralls> OK.
18:15:40 <hptina> * 00:14:53 WARN <gnc.gui> [gnc_load_app_icons()] No icon named 'gnucash-icon' found. Some gui elements may be missing their icons
18:15:40 <hptina> * 00:14:53 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_bool()] Unable to set value for key show-at-startup in schema dialogs.totd
18:15:40 <hptina> * 00:14:53 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_int()] Unable to set value for key current-tip in schema dialogs.totd
18:15:40 <hptina> * 00:14:54 OTHER <GLib-GIO> Settings schema 'org.gnucash.GnuCash.dialogs.import.hbci' is not installed
18:16:49 <jralls> rm /usr/local/share/glib-2.0/schemas/org.gnucash* && make install
18:18:49 *** frakturfreak1 has quit IRC
18:19:04 *** frakturfreak1 has joined #gnucash
18:21:02 <jralls> hptina did that not make sense?
18:21:45 <hptina> no, I don't think so ;)
18:22:59 <jralls> Run that. The first part clears out the installed org.gnucash schemas -- make uninstall seems not to--and the second re-installs GnuCash, I hope with the new schemas.
18:23:35 <hptina> yes, but it crashes again :(
18:24:10 <jralls> with the same import.hbci message in gnucash.trace?
18:24:38 <hptina> I'll take a look ...
18:25:08 <hptina> * 00:21:05 WARN <gnc.gui> [gnc_load_app_icons()] No icon named 'gnucash-icon' found. Some gui elements may be missing their icons
18:25:08 <hptina> * 00:21:05 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_bool()] Unable to set value for key show-at-startup in schema dialogs.totd
18:25:08 <hptina> * 00:21:05 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_int()] Unable to set value for key current-tip in schema dialogs.totd
18:25:08 <hptina> * 00:21:06 OTHER <GLib-GIO> Settings schema 'org.gnucash.GnuCash.dialogs.import.hbci' is not installed
18:26:05 <hptina> absolutely the same ... :(
18:26:05 <jralls> Hmm. Just for drill make sure that /usr/share/glib-2.0/schemas doesn't have any org.gnucash ones listed.
18:27:38 <jralls> what's absolutely the same?
18:27:56 <hptina> ls -l /usr/share/glib-2.0/schemas/*cash* <-- file not found
18:28:18 <hptina> the same output than before
18:31:20 <jralls> Right. So make uninstall, remove the org.gnucash.* schemas from /usr/local/share, rm -rf build-dir, and re-run cmake, make, and make install.
18:35:19 <hptina> ok, but I'll reboot before that
18:36:20 *** hptina has quit IRC
18:39:58 <jralls> @tell hptina I have to go afk for an hour or so.
18:39:58 <gncbot> jralls: The operation succeeded.
18:40:07 *** hptina has joined #gnucash
18:49:06 *** ChanServ sets mode: +v hptina
18:49:38 <hptina> re :)
18:49:38 <gncbot> hptina: Sent 9 minutes ago: <jralls> I have to go afk for an hour or so.
18:58:37 <hptina> * 00:57:44 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_bool()] Unable to set value for key show-at-startup in schema dialogs.totd
18:58:38 <hptina> * 00:57:44 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_int()] Unable to set value for key current-tip in schema dialogs.totd
18:58:38 <hptina> * 00:57:45 WARN <gnc.app-utils.gsettings> [gnc_gsettings_reset()] Resetting key reg-is-acct-pay-rec
18:59:45 <hptina> atm I'm running the latest 5.4.x kernel from Ubuntu
19:04:01 *** fell has quit IRC
19:09:02 *** fell has joined #gnucash
19:09:02 *** ChanServ sets mode: +o fell
19:20:57 *** lmat has quit IRC
19:24:32 <hptina> and here is the output from strace: https://gist.github.com/hptina-bm/820bf64fa4e7d2d60bec7d9e0b902ad1
19:30:12 *** reactormonk[m] has quit IRC
19:30:14 *** peter-butler[m] has quit IRC
19:30:19 *** dtux[m] has quit IRC
19:30:21 *** PeterScholtens[m] has quit IRC
19:30:32 *** ErikSchillingAblu[m] has quit IRC
19:30:33 *** Julianold[m] has quit IRC
19:30:43 *** luwum[m] has quit IRC
19:30:45 *** cybertani[m] has quit IRC
19:30:46 *** Trygve[m]1 has quit IRC
19:30:47 *** shukon[m] has quit IRC
19:30:49 *** technicus[m] has quit IRC
19:30:50 *** ElonSatoshi[m] has quit IRC
19:30:51 *** antonlada[m] has quit IRC
19:30:54 *** Amanda[m] has quit IRC
19:30:56 *** MatrixTravelerbot[m] has quit IRC
19:31:31 *** chris has joined #gnucash
19:31:31 *** ChanServ sets mode: +v chris
19:31:31 *** gncbot sets mode: +o chris
19:34:00 <hptina> full file: https://gist.githubusercontent.com/hptina-bm/820bf64fa4e7d2d60bec7d9e0b902ad1/raw/79831e3b36d8d19ddc79b788ae8e5371f50ca2a5/gistfile1.txt
19:43:26 <jralls> Pi0b@1reachD
19:44:40 <hptina> ???
19:45:19 <jralls> sorry, wrong window.
19:45:29 <hptina> :D
19:46:23 <jralls> strace output isn't terribly helpful in this case.
19:46:40 <hptina> :(
19:48:21 <hptina> openat(AT_FDCWD, "/var/lib/flatpak/exports/share/glib-2.0/schemas/gschemas.compiled", O_RDONLY) = -1 ENOENT (No such file or directory)
19:48:21 <hptina> openat(AT_FDCWD, "/home/tina/.local/share/flatpak/exports/share/glib-2.0/schemas/gschemas.compiled", O_RDONLY) = -1 ENOENT (No such file or directory)
19:48:21 <hptina> openat(AT_FDCWD, "/usr/share/plasma/glib-2.0/schemas/gschemas.compiled", O_RDONLY) = -1 ENOENT (No such file or directory)
19:48:21 <hptina> openat(AT_FDCWD, "/home/tina/.local/share/glib-2.0/schemas/gschemas.compiled", O_RDONLY) = -1 ENOENT (No such file or directory)
19:48:40 <hptina> isn't that helpful?
19:49:55 <jralls> Well it's interesting. Is the last line true?
19:50:42 <hptina> one moment, please
19:54:13 <hptina> yes, that's from strace's output
19:55:56 <jralls> I meant does the file really not exist, but I realize that I misread the path, so it shouldn't exist.
19:57:09 <hptina> after that II made a symlink, so the output is now:
19:57:14 <jralls> Just to be safe though, /usr/local/share/glib-2.0/schemas/gschemas.compiled does exist, right? If so then the problem may be that gsettings isn't looking in the right place for gschemas.compiled.
19:57:18 <hptina> *I
19:57:28 <hptina> openat(AT_FDCWD, "/var/lib/flatpak/exports/share/glib-2.0/schemas/gschemas.compiled", O_RDONLY) = -1 ENOENT (No such file or directory)
19:57:28 <hptina> openat(AT_FDCWD, "/home/tina/.local/share/flatpak/exports/share/glib-2.0/schemas/gschemas.compiled", O_RDONLY) = -1 ENOENT (No such file or directory)
19:57:28 <hptina> openat(AT_FDCWD, "/usr/share/plasma/glib-2.0/schemas/gschemas.compiled", O_RDONLY) = -1 ENOENT (No such file or directory)
19:57:28 <hptina> openat(AT_FDCWD, "/home/tina/.local/share/glib-2.0/schemas/gschemas.compiled", O_RDONLY) = 12
19:58:15 <hptina> yes, that file exists
19:59:06 <hptina> now that symlink points to "/home/tina/.local/share/glib-2.0/schemas/gschemas.compiled"
19:59:15 *** ChanServ sets mode: +v chf1
19:59:32 *** chris has quit IRC
19:59:33 <chf1> Mine is in "/usr/share/glib-2.0/schemas/gschemas.compiled".
19:59:55 <hptina> but gnucash still crashes
20:00:31 <jralls> That's for the installed debs. Does $XDG_DATA_DIRS include /usr/local/share?
20:00:48 <chf1> When? Immediately after startup?
20:01:16 <jralls> Right now. Do echo $XDG_DATA_DIRS and see if /usr/local/share is in the list.
20:01:21 <hptina> echo $XDG_DATA_DIRS
20:01:30 *** chris has joined #gnucash
20:01:30 *** ChanServ sets mode: +v chris
20:01:35 *** gncbot sets mode: +o chris
20:01:47 <hptina> /usr/share/plasma:/home/tina/.local/share/flatpak/exports/share:/var/lib/flatpak/exports/share:/usr/local/share:/usr/share
20:01:55 <chf1> Oh, yes, for flatpaks it is at different locations, one for each "instance".
20:02:36 <chf1> Example: /var/lib/flatpak/app/org.gnucash.GnuCash/x86_64/master-C3.8b-532-g6b55222ca-D3.8b-58-g3191b32/698d598b8a09a0502addd5f261b884a2a31c381223ee5507d4bfe6b66f049f06/files/share/glib-2.0/schemas/gschemas.compiled
20:02:45 <jralls> chf1, please study the whole conversation before adding noise.
20:03:01 <hptina> chf1, well, I don't use a flatpak, I built from source
20:03:57 <chf1> I thought I did so, jralls, but missed the beginning, sorry.
20:04:30 <hptina> openat(AT_FDCWD, "/run/dconf/user/1001", O_RDONLY) = -1 ENOENT (No such file or directory)
20:04:30 <hptina> openat(AT_FDCWD, "/run/user/1001/dconf/profile", O_RDONLY) = -1 ENOENT (No such file or directory)
20:04:30 <hptina> openat(AT_FDCWD, "/etc/dconf/profile/user", O_RDONLY) = -1 ENOENT (No such file or directory)
20:04:30 <hptina> openat(AT_FDCWD, "/usr/share/plasma/dconf/profile/user", O_RDONLY) = -1 ENOENT (No such file or directory)
20:04:30 <hptina> openat(AT_FDCWD, "/home/tina/.local/share/flatpak/exports/share/dconf/profile/user", O_RDONLY) = -1 ENOENT (No such file or directory)
20:04:31 <hptina> openat(AT_FDCWD, "/var/lib/flatpak/exports/share/dconf/profile/user", O_RDONLY) = 12
20:04:57 <hptina> here the flatpak file is found ...
20:05:23 <jralls> So you do use flatpak, just not for GnuCash.
20:05:41 <hptina> could that be the reason for that issue?
20:06:36 <hptina> jralls, no, but there is a dconf file in a flatpak path
20:06:45 <jralls> You shouldn't include flatpak directories in any paths used outside of that particular flatpak, and in general you should leave it to flatpak run to set up the environment inside the flatpak.
20:07:37 <hptina> openat(AT_FDCWD, "/var/lib/flatpak/exports/share/dconf/profile/user", O_RDONLY) = 12 <-- this file gets found and opened automatically
20:07:49 <jralls> And that's bad.
20:08:08 <hptina> I think so, too
20:08:44 <jralls> If you don't use flatpak for anything you should uninstall flatpak and rm -rf /var/lib/flatpak.
20:08:46 <hptina> let's make another symlink ...
20:09:11 <jralls> Not until you clean out all of the flatpak leftovers.
20:09:59 <jralls> Next, what's /usr/share/plasma about? What's plasma?
20:12:07 <chf1> That "flatpak" in the path lists above was the thing which misleaded me in the first place… I don't have anything in "/var/lib/flatpak/exports/" except ".desktop" files and icons despite actually using flatpak.
20:12:24 <chf1> Plasma is the KDE desktop stuff.
20:13:51 <chf1> Contains "/usr/share/plasma/shells/org.kde.plasma.desktop/contents/updates/desktop-base.js" ONLY here (KDE4.x).
20:14:11 <jralls> OK. Try export XDG_DATA_DIRS=/usr/local/share:/usr/share and then run GnuCash.
20:15:39 <hptina> next step:
20:15:46 <hptina> openat(AT_FDCWD, "/run/dconf/user/1001", O_RDONLY) = -1 ENOENT (No such file or directory)
20:15:46 <hptina> openat(AT_FDCWD, "/run/user/1001/dconf/profile", O_RDONLY) = -1 ENOENT (No such file or directory)
20:15:46 <hptina> openat(AT_FDCWD, "/etc/dconf/profile/user", O_RDONLY) = 12
20:16:01 <hptina> ok, let me try that
20:16:43 <jralls> Um, is dconf-service running?
20:17:24 <hptina> * 02:16:43 WARN <gnc.gui> [gnc_load_app_icons()] No icon named 'gnucash-icon' found. Some gui elements may be missing their icons
20:17:25 <hptina> * 02:16:43 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_bool()] Unable to set value for key show-at-startup in schema dialogs.totd
20:17:25 <hptina> * 02:16:43 ERROR <gnc.app-utils.gsettings> [gnc_gsettings_set_int()] Unable to set value for key current-tip in schema dialogs.totd
20:17:25 <hptina> * 02:16:43 OTHER <GLib-GIO> Settings schema 'org.gnucash.GnuCash.dialogs.import.hbci' is not installed
20:18:17 *** peter-butler[m] has joined #gnucash
20:18:46 <hptina> dconf-service: there is no such binary
20:19:07 <chf1> /usr/lib/dconf/dconf-service
20:19:37 <jralls> OK, the original crash. Progress, I guess. Is /usr/local/share/glib-2.0/schemas/gschemas.compiled older than org.gnucash.* ?
20:19:38 *** mechovation has joined #gnucash
20:19:38 <chf1> It's not in /usr/bin
20:20:10 <hptina> sorry, I found it, but it's not running
20:20:25 *** mechovation has left #gnucash
20:20:39 <jralls> That would be a good reason that dconf calls crash.
20:21:51 <hptina> gschemas.compiled is younger
20:22:57 <jralls> Hmm. Does strings .../gschemas.compiled | grep dialogs.import.hbci find one entry or two?
20:23:40 <hptina> from the manpage: "Users or administrators should never need to start the service, as it will be automatically
20:23:40 <hptina> started by dbus-daemon(1) whenever an application tries to write settings."
20:25:00 *** dtux[m] has joined #gnucash
20:25:23 <jralls> Oh, and you didn't actually check if it's running, either, did you?
20:25:35 <jralls> As in ps -A | grep dconf
20:26:13 <hptina> org.gnucash.dialogs.import.hbci
20:26:13 <hptina> org.gnucash.dialogs.import.hbci.connection-dialog
20:26:13 <hptina> (as(y(ys)))org.gnucash.dialogs.import.hbci.connection-dialog
20:27:30 <hptina> no, that service is not running
20:30:08 <jralls> That's suspicious.
20:31:08 *** guak has quit IRC
20:31:09 <jralls> But it also seems that you've lost gjanssens patch, you've got a truncated gschemas.compiled.
20:31:49 *** ErikSchillingAblu[m] has joined #gnucash
20:32:18 <hptina> openat(AT_FDCWD, "/etc/dconf/profile/user", O_RDONLY) = 12
20:32:18 <hptina> fstat(12, {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0
20:32:18 <hptina> read(12, 0x55995b2cf410, 4096) = -1 EISDIR (Is a directory)
20:32:18 <hptina> close(12) = 0 <-- does that matter?
20:32:21 <jralls> That strings..grep should return 10 lines including org.gnucash.dialogs.import.hbci.connection-dialog and org.gnucash.GnuCash.dialogs.import.hbci.connection-dialog.
20:34:02 <jralls> No, that just means that it read the file into memory and closed it when done.
20:34:33 <hptina> no, the patch is still applied: diff ../libgnucash/app-utils/gnc-gsettings.cpp ../libgnucash/app-utils/gnc-gsettings.cpp.orig.bak
20:34:34 <hptina> 558,559d557
20:34:34 <hptina> < {
20:34:34 <hptina> < PWARN("Resetting key %s", key);
20:34:34 <hptina> 561d558
20:34:36 <hptina> < }
20:34:37 <jralls> The line above matters: It means that something that was supposed to be a file is a directory.
20:34:49 <hptina> oh, no!
20:34:57 *** luwum[m] has joined #gnucash
20:35:05 <hptina> that's the wrong patch
20:35:55 <jralls> Wrong patch. The one that's important is in gnucash/gschemas/org.gnucash.GnuCash.deprecated.gschema.xml.in.
20:35:56 *** cybertani[m] has joined #gnucash
20:36:27 <jralls> See https://github.com/Gnucash/gnucash/commit/b7001b52cf2842f056b7ad919791734462f2b1a4 for the diff.
20:36:48 *** shukon[m] has joined #gnucash
20:37:34 *** technicus[m] has joined #gnucash
20:37:49 *** ElonSatoshi[m] has joined #gnucash
20:38:26 <jralls> But the inability to read /etc/dconf/profile/user might be what's causing dconf-server to shut down.
20:38:27 <hptina> grep '<child name="connection-dialog" schema="org.gnucash.dialogs.import.hbci.connection-dialog"/>' ../gnucash/gschemas/org.gnucash.GnuCash.deprecated.gschema.xml.in
20:38:27 <hptina> <child name="connection-dialog" schema="org.gnucash.dialogs.import.hbci.connection-dialog"/>
20:38:46 <hptina> so the patch is applied
20:39:14 <jralls> So why do you intermittently get the wrong schemas?
20:39:57 <hptina> should /etc/dconf/profile/user be a directory?
20:40:49 <jralls> No, it should be a text file.
20:41:17 <jralls> dconf-server tried to read it and got an error saying that it's a directory.
20:41:37 <jralls> And I guess that made dconf-server quit.
20:41:50 <hptina> hmm, I symlinked it to a directory
20:42:05 <hptina> sorry, my fault
20:42:37 <chf1> I don't have "/etc/dconf/profile/user" at all, and dconf is running (perhaps a matter of version, or it's optional). Is "/usr/bin/dbus-daemon" running? This should start the service if needed.
20:44:02 <hptina> because /var/lib/flatpak/exports/share/dconf/profile/user is a directory, too
20:45:05 <jralls> I'm leaning toward you need to move your data files to an external disk and reformat and reinstall Ubuntu from the ISO.
20:45:11 <hptina> message+ 891 0.4 0.0 8968 6124 ? Ss 00:37 0:32 /usr/bin/dbus-daemon --system --address=systemd: --nofork --nopidfile --systemd-activation --syslog-only
20:45:11 <hptina> tina 1630 0.0 0.0 8040 5304 ? Ss 00:37 0:02 /usr/bin/dbus-daemon --session --address=systemd: --nofork --nopidfile --systemd-activation --syslog-only
20:45:11 <hptina> tina 1906 0.0 0.3 233144 24060 ? Sl 00:37 0:00 /usr/bin/gmenudbusmenuproxy
20:45:11 <hptina> tina 2031 0.0 0.0 7240 4036 ? S 00:37 0:00 /usr/bin/dbus-daemon --config-file=/usr/share/defaults/at-spi2/accessibility.conf --nofork --print-address 3
20:45:40 <chf1> looks correct.
20:46:49 *** antonlada[m] has joined #gnucash
20:46:56 *** sergiomiguelrp has quit IRC
20:47:08 <chf1> I wonder where /var/lib/flatpak/exports/share/dconf/profile/user comes from. I don't have it (several flatpaks insalled).
20:48:15 <chf1> ls -l /var/lib/flatpak/exports/share/
20:48:15 <chf1> insgesamt 8
20:48:15 <chf1> drwxrwxr-x 2 root root 4096 Sep 28 00:51 applications
20:48:15 <chf1> drwxrwxr-x 3 root root 4096 Sep 22 2019 icons
20:50:41 <chf1> I.e. only start menu entries / desktop icons in the "exports" directory.
20:52:24 <hptina> ls -l /var/lib/flatpak/exports/share/
20:52:25 <hptina> total 16
20:52:25 <hptina> drwxr-xr-x 2 root root 4096 Sep 24 20:46 applications
20:52:25 <hptina> drwxr-xr-x 3 root root 4096 Feb 22 2021 dbus-1
20:52:25 <hptina> drwxr-xr-x 3 root root 4096 Mar 3 2021 dconf
20:52:25 <hptina> drwxr-xr-x 3 root root 4096 Feb 22 2021 icons
20:53:07 <hptina> dconf exists since March
20:53:59 <chf1> Do you remember what you've done with flatpaks then?
20:54:37 <hptina> I don't work often with flatpaks
20:55:15 <hptina> but I don't remember any details
20:55:27 <chf1> I remember having used a configruation application once. You could change settings there, for example which files/directories are allowed to be accessed from inside a flatpak sandbox.
20:56:37 <hptina> no, I didn't use any app like that (AFAIR)
20:56:48 <chf1> Bit this should work in the other direction only, those flatpak specific files are in places not normally in any path, thus should not interfere with isntalled normal packages, or locally compiled stuff.
20:56:53 <chf1> But
20:57:27 <ArtGravity> Is there a way to run an expense report that only shows expenses related to another account?
20:58:00 <ArtGravity> e.g. All expense accounts that have a specific bank/credit card, etc as the corresponding account
20:58:11 <hptina> openat(AT_FDCWD, "/var/lib/flatpak/exports/share/dconf/profile/user", O_RDONLY) = 12
20:58:11 <hptina> fstat(12, {st_mode=S_IFDIR|0777, st_size=4096, ...}) = 0
20:58:11 <hptina> read(12, 0x55de4f375c00, 4096) = -1 EISDIR (Is a directory) <-- but not on my system ...
20:59:18 <hptina> well, my bed is waiting for me
20:59:25 <warlord> hptina, ls -ld /var/lib/flatpak/exports/share/dconf/profile/user
21:00:00 <hptina> good night, folks - thank you very much so far! :))
21:00:59 <hptina> drwxrwxrwx 2 root root 4096 Mar 4 2021 /var/lib/flatpak/exports/share/dconf/profile/user
21:01:26 <hptina> warlord, and now?
21:02:07 <warlord> hptina, that's a directoyy
21:02:10 <chf1> I've found something, hptina, those rhings under "exports" seem to be meant to be included into your normal system environment.
21:02:12 <warlord> see that 'd' in there?
21:02:19 <warlord> drwxrwxrwx means it's a directory
21:02:21 <chf1> things
21:02:38 <hptina> warlord, yes, I already knew that ;)
21:02:56 <chf1> So I would try to RENAME it, or move it out of the way.
21:03:34 <warlord> hptina, then what's the question? That's why read() is returning EISDIR. It's trying to open a directory and read it like a file.
21:03:42 <hptina> chf1, renaming it seems to be a good idea
21:04:40 <chf1> Like so: mkdir /var/lib/flatpak/exports/share/tmp ; mv /var/lib/flatpak/exports/share/d* /var/lib/flatpak/exports/share/tmp
21:05:12 <chf1> i.e. move the dbus/dconf subdirs into "tmp".
21:05:28 *** chf1 is now known as chf
21:05:55 <warlord> anyways, i'm off for the night. TTFN
21:06:04 <chf> Good night
21:06:10 <jralls> G'night, warlord.
21:06:26 <jralls> It'll be pumpkin time for me to shortly.
21:06:41 <hptina> now the error message is smaller:
21:06:46 <hptina> * 03:05:35 WARN <gnc.gui> [gnc_load_app_icons()] No icon named 'gnucash-icon' found. Some gui elements may be missing their icons
21:06:46 <hptina> * 03:05:35 OTHER <GLib-GIO> Settings schema 'org.gnucash.GnuCash.dialogs.import.hbci' is not installed
21:07:20 <hptina> but gnucash still crashes
21:07:53 <jralls> Well, yes, that's the sigtrap crash that gjanssens's patch is supposed to fix.
21:08:26 <jralls> And it's caused by a truncated gschemas.compiled.
21:09:05 <jralls> You might try glib-compile-gschemas /usr/local/share/glib-2.0/schemas
21:09:54 <jralls> Without the --strict arg in used in cmake it might get you a working gschemas.compiled
21:10:16 <hptina> mkdir("/run/user/1001/dconf", 0700) = -1 EEXIST (File exists)
21:10:17 <hptina> stat("/run/user/1001/dconf", {st_mode=S_IFDIR|0700, st_size=80, ...}) = 0
21:10:17 <hptina> openat(AT_FDCWD, "/run/user/1001/dconf/user", O_RDWR|O_CREAT, 0600) = 12
21:10:17 <hptina> pwrite64(12, "\0", 1, 1) = 1
21:10:17 <hptina> mmap(NULL, 1, PROT_READ, MAP_SHARED, 12, 0) = 0x7fdfe2328000
21:10:17 <hptina> close(12) = 0
21:10:19 <chf> Or it gets overridden by a file found first in the paths not containing that patch?
21:10:19 <hptina> openat(AT_FDCWD, "/home/tina/.config/dconf/user", O_RDONLY) = 12
21:10:21 <hptina> fstat(12, {st_mode=S_IFREG|0664, st_size=6281, ...}) = 0
21:10:23 <hptina> mmap(NULL, 6281, PROT_READ, MAP_PRIVATE, 12, 0) = 0x7fdfd67a4000
21:10:25 <hptina> close(12)
21:13:36 <hptina> jralls, it's up and running! :))
21:14:06 <hptina> jralls> You might try glib-compile-gschemas /usr/local/share/glib-2.0/schemas <-- this one helped!
21:14:11 <chf> Solution? What jralls suggested, hptina?
21:14:23 <chf> Ah, perfect.
21:14:23 <jralls> Workaround, not a solution.
21:17:28 <hptina> jralls, can't thiss workaround not be integrated in an install script?
21:17:39 <hptina> *this
21:18:54 <jralls> It would be better not to. It would be better still to wire it so that it stops ninja instead of writing a defective gschemas.compile and moving on.
21:19:56 <hptina> maybe I should compile with meson/ninja next time?
21:21:38 <jralls> Oh, that's an interesting thought. It would be cmake/ninja, but yeah, that might be part of the problem. gjanssens and I always use ninja so we might not see a make problem. I'll check that now.
21:22:28 <chf> If this consistently works for you, you can also make your own wrapper build script calling this after the build process.
21:23:40 *** Amanda[m] has joined #gnucash
21:24:07 <hptina> glib-compile-schemas --strict /usr/local/share/glib-2.0/schemas <-- this command works, too
21:24:32 <chf> By the way, is there a reason not to use the Ubuntu Gnucash package, hptina?
21:25:23 <hptina> chf, the Ubuntu deb-package is too old (IMHO)
21:27:02 <jralls> hptina It works with --strict? That's interesting and supports the idea that something's wrong with make install.
21:27:03 <chf> Same here, but only because I'm using (old?)oldstable which doesn't have FinTS with PSD2. (German online banking)
21:27:04 <hptina> I want to use gnucash as a replacement for moneyplex (I use HBCI/FinTS with a chipcard)
21:27:54 <chf> Works fine here with one of the current Flatpaks.
21:28:30 <hptina> jralls, yes, the resulting file (gschemas.compiled) has the same size as before
21:29:07 * hptina doesn't really like flatpaks ...
21:31:55 <chf> I think I've also rather outdated libs, so compiling would be difficult as well.
21:32:37 <hptina> jralls> I'm leaning toward you need to move your data files to an external disk and reformat and reinstall Ubuntu from the ISO. <-- NO!!!
21:33:18 <hptina> chf, which distro are you running?
21:33:57 <chf> Debian, very old, but with a few exceptions, everything works fine.
21:35:45 <hptina> chf, can't you upgrade your Debian to the latest stable version without reinstalling it?
21:35:53 <chf> The exceptions are those things which "get too old" due to external influences, like browsers getting rejected by stupid web designers, and Gnucash failing because of new HBCI TAN methods.
21:36:55 <hptina> chf, you should always use a recent browser!
21:36:56 <chf> I could, but that would de-install a lot, and there's anothe difficulty: this was once an i386 intsllation.
21:37:51 *** ArtGravity has quit IRC
21:37:55 <hptina> chf, oh, then you seem to have a really old machine there ;)
21:38:21 <chf> Pff, that's Linux, not Windows, so a non-recent browser is a lesser problem as you would normally expect.
21:38:55 <chf> It is now a multi-arch system, but "the other way around".
21:39:20 <hptina> chf, Linux has its security flaws, too ...
21:39:52 <chf> Most things are AMD64, but dpkg/apt "think" of it as i586.
21:40:21 <hptina> chf, somewhat strange ;)
21:40:31 <chf> Of course, but on Linux the main problem is spammers trying to break into your web server.
21:40:56 <hptina> yes, surely!
21:41:36 <chf> They normally don't try to take over your client for botnet integration, this is mainly a Windows problem.
21:42:51 <chf> So my internet server is up to date and secure, and my clients are not, which in practice doesn't matter, provided you aren't really stupid.
21:43:05 <hptina> chf, and an IoT problem, too ..
21:46:09 <chf> I avoid most IoT things, especially those where the vendor is "root" rather than me.
21:46:43 *** MatrixTravelerbot[m] has joined #gnucash
21:47:01 <chf> Those phoning-home-smartphone-cloud-devices…
21:48:57 <hptina> *UGLY*
21:50:26 <hptina> so, good night, folks an thank you very much for helping me! :))
21:50:35 <chf> Good night!
21:53:32 *** hptina has quit IRC
22:13:37 *** joo has quit IRC
22:57:29 <jralls> @tell gjanssens hptina suggested that there might be an issue with make vs. ninja install after they successfully ran glib-compile-schemas --strict /usr/local/share/glib-2.0/schemas. I tested with make and indeed gschemas.compiled containes only the old paths.
22:57:29 <gncbot> jralls: The operation succeeded.
23:03:39 *** fell has quit IRC
23:03:59 *** fell has joined #gnucash
23:04:00 *** ChanServ sets mode: +o fell
23:28:27 *** reactormonk[m] has joined #gnucash
23:29:19 *** fell has quit IRC
23:29:39 *** fell has joined #gnucash
23:29:39 *** ChanServ sets mode: +o fell