2013-10-07 GnuCash IRC logs
00:00:29 *** fell_ has quit IRC
00:03:15 *** StuM has quit IRC
00:05:42 *** StuM2 has quit IRC
00:35:48 *** blathijs has quit IRC
00:36:16 *** blathijs has joined #gnucash
01:39:48 *** Jimraehl1 has quit IRC
02:03:56 *** ErKa has quit IRC
03:39:46 *** Ard0nik has joined #gnucash
03:48:23 *** Ardonik has quit IRC
04:10:00 *** john has quit IRC
05:16:41 *** Ard0nik has quit IRC
06:32:00 *** nomeata has joined #gnucash
06:36:32 *** fell_ has joined #gnucash
06:36:33 *** gncbot sets mode: +o fell_
06:37:12 *** Jimraehl1 has joined #gnucash
06:56:13 *** Jimraehl1 has left #gnucash
06:58:34 *** Jimraehl1 has joined #gnucash
07:02:14 *** stl has quit IRC
07:31:16 *** gjanssens-afk is now known as gjanssens
07:39:50 *** stl has joined #gnucash
09:05:02 *** Topcat has joined #gnucash
09:30:01 *** benoitg has quit IRC
09:35:28 *** Topcat has quit IRC
09:35:50 *** Topcat has joined #gnucash
09:51:23 *** Topcat has quit IRC
09:51:46 *** Topcat has joined #gnucash
09:56:45 *** Topcat has quit IRC
09:57:33 *** Topcat has joined #gnucash
10:01:39 *** rpg has quit IRC
10:07:26 *** benoitg has joined #gnucash
10:12:46 *** Topcat has quit IRC
10:13:09 *** Topcat has joined #gnucash
10:15:40 *** ErKa has joined #gnucash
10:23:13 *** Topcat has quit IRC
10:23:36 *** Topcat has joined #gnucash
10:30:40 *** Topcat has quit IRC
10:31:17 *** Topcat has joined #gnucash
10:34:16 *** john has joined #gnucash
10:34:17 *** gncbot sets mode: +o john
10:46:27 *** nomeata has quit IRC
10:48:27 *** aqua___ has joined #gnucash
10:54:36 *** aqua___ has quit IRC
10:56:56 *** benoitg has quit IRC
10:56:57 *** aqua___ has joined #gnucash
10:58:41 <gjanssens> I have pushed my gsettings branch to trunk
11:01:07 <gjanssens> I was waiting for svn to send the commit messages, but the last message I get is for r23241
11:01:17 <gjanssens> The last commit is r23273 though
11:01:49 <gjanssens> And while I write another set of mails is coming ing
11:01:54 <gjanssens> s/ing/in/
11:02:13 <gjanssens> Apparently mail processing takes some time..
11:03:17 <gjanssens> The full patch set is in svn trunk though, so it's probably fine
11:05:14 *** linas_ has quit IRC
11:07:56 *** benoitg has joined #gnucash
11:13:29 *** Topcat has quit IRC
11:13:53 *** Topcat has joined #gnucash
11:14:15 <gjanssens> It's the biggest patch set I have done so far on GnuCash
11:14:56 <gjanssens> With changes throughout all of the gui code
11:15:19 <gjanssens> I hope it won't break too many patches still lingering in bugzilla
11:15:31 <gjanssens> Or in other dev's local checkouts :(
11:17:20 *** aqua___ has quit IRC
11:17:23 <gjanssens> john, jhbuild build wants to rebuild each and every package again...
11:17:45 <gjanssens> and fails at step gnome-doc-utils on my system
11:18:01 <gjanssens> I already tried "wipe directory and start over"
11:18:17 <gjanssens> but it gives
11:18:19 <gjanssens> jhbuild build: failed to unpack /Users/janssege/gtk/source/pkgs/gnome-doc-utils-0.20.10.tar.xz
11:19:17 <gjanssens> And something else:
11:19:48 <gjanssens> mails from our git mail script may have an issue as well
11:20:11 <gjanssens> my large patch set only resulted in one mail from the git mail script
11:20:37 <gjanssens> this mail only mentions 3 of the nearly 70 commits I pushed
11:26:12 *** aqua___ has joined #gnucash
11:26:55 <warlord> gjanssens: I see mail about 23273
11:27:09 <warlord> BIAB
11:27:18 *** warlord has quit IRC
11:30:10 <mikee> gjanssens: Wow! Compiling it now.
11:31:15 <gjanssens> @tell warlord The last one came in here two minutes ago
11:31:15 <gncbot> gjanssens: The operation succeeded.
11:32:13 <gjanssens> mikee: :)
11:34:16 *** Topcat has quit IRC
11:34:55 *** benoitg has quit IRC
11:38:24 <john> Geert.
11:38:41 <john> Make distcheck fails with
11:38:42 <john> make[6]: Entering directory `/home/john/gnucash-shag/gnucash-2.5.5/_build/src/business/business-gnome/gschemas'
11:38:43 <john> make[6]: *** No rule to make target `org.gnucash.dialogs.business.gschema.xml', needed by `org.gnucash.dialogs.business.gschema.valid'. Stop.
11:38:44 <john> m
11:41:50 <john> gjanssens: Geert, faling to unpack a tar.xz suggests that you forgot to bootstrap your build, since xz is in bootstrap. Run jhbuild bootstrap, then try the build again.
11:50:52 <gjanssens> john: I just pushed a change to fix the dist check error
11:51:05 <gjanssens> I also updated POTFILES.in
11:51:08 <mikee> Successful compile on Fedora18
11:51:19 <gjanssens> mikee: cool !
11:51:25 <gjanssens> Let's hope it runs also
11:51:55 <gjanssens> john: regarding jhbuild, in what circumstances do I have to rerun it ?
11:52:00 <mikee> Doing distcheck. I'll let you know ASAP/
11:52:13 <gjanssens> I didn't remove my jhbuild config
11:52:38 <gjanssens> I couldn't rebuild gnucash-git via jhbuild shell
11:52:50 <gjanssens> So I decided to run jhbuild build instead
11:53:33 <gjanssens> which got me in the situation that gnome-doc-utils wouldn't build anymore
11:53:40 <gjanssens> Puzzled...
11:54:17 <gjanssens> anyway, jhbuild bootstrap just finished and gnome-doc-utils is building again
11:55:07 <gjanssens> On gnucash-git you can now remove ORBit and GConf as dependencies by the way :)
11:55:09 <john> Geert, it's not the config. jhbuild bootstrap supplies the tools that Apple leaves out or doesn't supply recent-enough versions of. I've fiddled my build so that they go to ~/.local because I have a dozen different build trees for testing, but normally if you update jhbuild (with gtk-osx-build-setup.sh) or start a new build tree, you need to run jhbuild bootstrap as the first step.
11:55:49 <gjanssens> ok
11:55:53 <john> > remove ORBit and GConf: And *dbus*! Yay!
11:56:03 <gjanssens> LOL
11:56:54 <john> And the stupid link into /Library! And the sudo script to install it. Double-yay!
11:57:05 <gjanssens> On my F19 system make distcheck fails because makeinfo is missing
11:57:10 <gjanssens> is that to be expected ?
11:57:29 <gjanssens> Party :D
11:57:58 <mikee> Ah, fails distcheck looking for DBD modules. Which is fair enough since I haven't installed any.
11:59:23 <john> ISTR make info is used for some of the old documentation in docs or src/docs, which is in texinfo format.
12:00:20 <john> Anything not under Doxygen control is seriously obsolete and should probably be removed. But not today. ;-)
12:00:28 <gjanssens> :-
12:00:33 <gjanssens> :-)
12:00:49 <john> I was wondering about :-
12:00:59 *** ErKa has quit IRC
12:01:02 <gjanssens> slip of the fingers
12:01:04 <mikee> Copy all gconf files to tmp dir /home/mikee/.gnc-migration-tmp
12:01:08 <mikee> warning: failed to load external entity "/home/mikee/progs/gnucash-trunk/share/gnucash/dialogs-import-generic-matcher.xml"
12:01:12 <mikee> Backtrace:
12:01:16 <mikee> In current input:
12:01:21 <mikee> 1: 0* [run-migration]
12:01:24 <mikee> In /home/mikee/.gnc-migration-tmp/migrate-prefs-user.scm:
12:01:30 <mikee> 132: 1* [gnc-prefs-set-bool "org.gnucash.dialogs.find" "search-for-active-only"]
12:01:51 <mikee> migrate-prefs-user.scm:132:1: In procedure gnc-prefs-set-bool in expression (gnc-prefs-set-bool "org.gnucash.dialogs.find" "search-for-active-only"):
12:01:59 <mikee> migrate-prefs-user.scm:132:1: Wrong number of arguments to #<primitive-procedure gnc-prefs-set-bool>
12:02:05 <mikee> Crash
12:02:38 <gjanssens> mikee: first the warning
12:02:40 <gjanssens> you can ignore it
12:02:53 <gjanssens> This is a side effect of how libxslt works
12:03:04 <gjanssens> I didn't find a way to avoid it
12:03:20 <john> Make distcheck fails with
12:03:41 <john> Attempt to migrate your preferences from the old to the new system
12:03:41 <john> Clear previous tmp dir /home/john/.gnc-migration-tmp
12:03:41 <john> Copy all gconf files to tmp dir /home/john/.gnc-migration-tmp
12:03:41 <john> warning: failed to load external entity "/home/john/gnucash-shag/gnucash-2.5.5/_inst/share/gnucash/make-prefs-migration-script.xsl"
12:03:41 <john> error
12:03:41 <john> xsltParseStylesheetFile : cannot parse /home/john/gnucash-shag/gnucash-2.5.5/_inst/share/gnucash/make-prefs-migration-script.xsl
12:03:41 <john> warning: failed to load external entity "/home/john/gnucash-shag/gnucash-2.5.5/_inst/share/gnucash/migratable-prefs.xml"
12:04:34 <john> Backtrace:
12:04:35 <john> In current input:
12:04:37 <john> 1: 0* (use-modules (migrate-prefs-user))
12:04:38 <john> 1: 1 (eval-case (# # *unspecified*) (else #))
12:04:39 <john> 1: 2 (begin (process-use-modules (list (list #))) *unspecified*)
12:04:41 <john> ?: 3* [process-use-modules (((migrate-prefs-user)))]
12:04:42 <john> ?: 4 (let* ((interfaces #)) (call-with-deferred-observers (lambda () #)))
12:04:44 <john> ?: 5* [map #<procedure #f (mif-args)> (((migrate-prefs-user)))]
12:04:46 <john> ?: 6* [#<procedure #f (mif-args)> ((migrate-prefs-user))]
12:04:47 <john> ?: 7* (or (apply resolve-interface mif-args) (error "no such module" mif-args))
12:04:48 <john> ?: 8* [apply #<procedure resolve-interface (name . args)> ((migrate-prefs-user))]
12:04:50 <john> ?: 9 [resolve-interface (migrate-prefs-user)]
12:04:52 <john> ...
12:04:53 <john> ?: 10 (let* (# # # # ...) (and # #) (if # public-i #))
12:04:54 <john> ?: 11* (and (or (not module) (not public-i)) (error "no code for module" name))
12:04:56 <john> ?: 12 [error "no code for module" (migrate-prefs-user)]
12:04:58 <john> ...
12:04:59 <john> ?: 13 [scm-error misc-error #f ...]
12:05:01 <john> <unnamed port>: In procedure scm-error in expression (scm-error (quote misc-error) #f ...):
12:05:03 <john> <unnamed port>: no code for module (migrate-prefs-user)
12:05:15 <gjanssens> mikee: can you look in /home/mikee/.gnc-migration-tmp/migrate-prefs-user.scm
12:05:54 <mikee> Yep
12:05:55 <gjanssens> and copy the guile snippet that attempts to set the option search-for-active-only
12:06:24 <gjanssens> It's odd the value is omitted from that option
12:06:47 <mikee> (gnc-prefs-set-bool "org.gnucash.dialogs.find" "search-for-active-only" )
12:07:02 <mikee> Not all in one line in reality.
12:07:45 <mikee> Missing preference value ? There's a blank line in the file
12:08:09 <mikee> line 138
12:08:55 <gjanssens> The line number is different for each user
12:09:11 <gjanssens> The file is generated at runtime based on the gconf settings you have set up
12:09:43 <mikee> Ah. Just after the comment ; preference value .
12:10:09 <gjanssens> Can you look in /home/mikee/.gnc-migration-tmp/dialogs-find.xml
12:10:25 *** warlord has joined #gnucash
12:10:26 *** gncbot sets mode: +o warlord
12:10:38 <gjanssens> And look for an entity called search_for_active_only
12:10:45 <mikee> Looking.
12:10:45 <gjanssens> (notice the underscores here)
12:10:57 <gjanssens> Does it have a value set ?
12:11:33 <mikee> No.
12:11:39 <mikee> <entry name="search_for_active_only" mtime="1331814199" schema="/schemas/apps/gnucash/dialogs/search_for_active_only"/>
12:12:24 *** benoitg has joined #gnucash
12:13:07 <mikee> Using gconf-editor I see that value isn't set in the orginial config.
12:13:13 <gjanssens> Damn
12:13:20 <gjanssens> That means I'm in for some fun
12:13:34 <gjanssens> Gconf allows entries to exist without a value
12:13:50 <gjanssens> Which is still different from not having set an entry at all
12:13:56 <gjanssens> I'll work on it
12:14:46 <gjanssens> john: I wonder why make-prefs-migration-script.xsl isn't installed
12:15:04 <warlord> gjanssens: how did you add it to Makefile.am?
12:15:05 <gncbot> warlord: Sent 43 minutes ago: <gjanssens> The last one came in here two minutes ago
12:15:05 <gjanssens> same goes for migratable-prefs.xml
12:15:22 <gjanssens> prefsmigrationdir = $(GNC_SHAREDIR)
12:15:23 <gjanssens> prefsmigration_DATA = \
12:15:24 <gjanssens> make-prefs-migration-script.xsl \
12:15:25 <john> Geert: Because _nothing_ is installed during make-distcheck.
12:15:26 <gjanssens> migratable-prefs.xml
12:15:51 <gjanssens> even tough it speaks of an _inst directory ?
12:16:11 <warlord> distcheck doesn't install.
12:16:13 <john> Yes.
12:16:35 <gjanssens> Ok, that reveals another issue I have to fix
12:16:39 <warlord> it runs, effectively: make dist; (unpackage new tarball and in it run:) configure, make, make check
12:17:11 <gjanssens> I chose to run the migration code at module load time of the app-utils module
12:17:20 <gjanssens> I realize now that's a bad idea
12:17:39 <gjanssens> that module is also loaded during testing, without it being installed
12:17:51 <gjanssens> I'll move the call
12:17:54 <gjanssens> BRB
12:19:14 <warlord> Why not put it in main() ?
12:22:31 <john> Needs Guile, so inner_main().
12:24:25 *** linas has joined #gnucash
12:25:28 <john> Not sure that that's going to fix the problem, though. It's going to have to take a path to find its script depending on whether it's installed or not.
12:27:49 <gjanssens> The migration should not run during make check IMO
12:28:11 <gjanssens> But yes, there are other situations gnucash is run without being installed
12:28:22 <gjanssens> I see this in my debugger for example
12:28:37 <gjanssens> (Using eclipse to develop)
12:29:07 <john> I run gnucash from the build directory often.
12:29:50 <john> OTOH, I don't need the migration script to run, I just need it to not crash.
12:30:29 *** Krzysiek_K has joined #gnucash
12:33:14 <warlord> modify the 'run-gnucash-from-build-tree script' then.. or base the file on the place where it would look for the 'environment' script.
12:33:32 <warlord> ... or just not error fatally if the script isnt found.
12:34:16 <john> I vote for the last option: Just check first to see if the script is present and move on if it isn't.
12:34:55 <john> But by all means move it out of instantiating an important module. Otherwise it's going to slow down running tests, and they already take too long.
12:35:23 <warlord> Yeah, inner_main() seems a good place to me.
12:36:01 *** jmd has joined #gnucash
12:36:28 <gjanssens> Yep, that's where it is now.
12:37:14 <gjanssens> I just have to add some code to skip the migration if the xsl and xml file are not found
12:37:18 <gjanssens> Then I'll push again
12:38:52 <mikee> Geert: If that's going to happen soon I'll hang around, otherwise I'll be AFK from a while. Tummy needs re-filling.
12:42:09 <warlord> mikee: with liquids or solids?
12:42:34 <mikee> Liquids obviously :)
12:42:53 <mikee> Brown and foamy liquids.
12:44:04 <john> So you're going to be a *lot* of help when you get back! ;-)
12:44:04 <gjanssens> mikee: go take care of your tummy :)
12:44:11 *** ErKa has joined #gnucash
12:44:48 <gjanssens> I am working on it, but don't know how long it will take really
12:45:28 <warlord> .... and /me just finished breakfast.
12:45:41 <mikee> How hard can typing "make install" be? Cheers have fun. I'll check in later.
12:46:14 <john> Derek, are you not at home? Should be lunchtime there...
12:47:45 <warlord> No, I'm in SF
13:01:58 *** benoitg has quit IRC
13:03:27 *** aqua___ has quit IRC
13:13:42 *** aqua___ has joined #gnucash
13:16:27 *** Topcat has joined #gnucash
13:17:37 *** ErKa has quit IRC
13:18:57 <gjanssens> john, mikee, two additional commits pushed
13:19:08 <gjanssens> which should fix the following issues
13:19:38 <gjanssens> move migration function call to inner_main, so it's not run at module load time anymore
13:20:18 <gjanssens> skip migration of the required xml/xsl files are not found (fixes crash when running from a not installed environment)
13:21:35 <gjanssens> skip preferences that are set in gconf, but without a value (should fix the search-for-active-only crash)
13:21:52 <gjanssens> Please retest ;)
13:22:10 * gjanssens is running a distcheck himself as well
13:22:31 <john> I've restarted distcheck on F18.
13:22:43 <john> Have you tried this on Windows?
13:23:12 <gjanssens> I tested the guile code independently of gnucash
13:23:25 <gjanssens> and I ran xsltproc to test the xsl transform
13:23:29 <gjanssens> both worked ok
13:23:55 <gjanssens> I have a windows build running, but I don't expect it to finish anytime soon...
13:24:24 <gjanssens> I can tell you multi-platform support is a bitch
13:24:55 <gjanssens> The times I had to redo my work because it worked fine on linux, but then failed on OS X or Windows...
13:25:22 <gjanssens> Guile behaves pretty differently on each platform
13:26:05 <gjanssens> Spaces in paths can bite as well (/Users/<user>/Library/Application Support/GnuCash anyone ?)
13:26:15 <john> I try to avoid working with guile. Makes me crazy regardless of platform.
13:26:23 <gjanssens> sure does
13:26:37 <warlord> Or C:\Program Files\...
13:26:48 <gjanssens> But it was the only option I had here barring keeping GConf around for another major release
13:27:05 <gjanssens> warlord: that one too...
13:27:26 <gjanssens> mingw wraps that fairly well, so in C you don't really notice
13:27:41 <gjanssens> But guile seems to have its own path parsing mechanism
13:27:52 <gjanssens> which trips over whitespace in filenames
13:33:45 <warlord> Yeah, guile..........
13:34:47 <john> Heh, you actually missed the distcheck error in src/business/business-gnome/gschemas/Makefile.am, which was (is, until I push) a typo in EXTRA_DIST. I guess you'll get to it in your own build.
13:37:00 <gjanssens> Yep, I got stuck there
13:37:40 <gjanssens> And looked too long at the code to spot it myself
13:40:20 <john> r23277
13:43:22 *** ErKa has joined #gnucash
13:43:34 <gjanssens> thanks
13:44:48 <gjanssens> I had fixed half of it btw in r23274. There were two bugs in that Makefile.am
13:44:58 <gjanssens> That's a lot of bugs for such a small file ;)
13:49:13 <gjanssens> aqbanking seems to have a concurrent build issue
13:49:36 <gjanssens> When running jhbuild, the first build attempt always fails
13:49:42 <gjanssens> A second attempt succeeds
13:49:44 <john> Aqbanking itself, or src/import-export/aqbanking?
13:49:59 <gjanssens> Aqbanking itself
13:50:15 <gjanssens> Not our concern really, but an observation
13:50:22 <john> Huh. Never noticed that.
13:50:30 <gjanssens> It's the second time I run into this
13:50:39 <gjanssens> It can be a matter of timing
13:51:07 <john> Yeah, and the small matter that Martin very much marches to his own drummer. :-/
13:56:52 *** Topcat has quit IRC
13:57:14 *** Topcat has joined #gnucash
13:57:23 <john> Distcheck completed. I'll proceed with preparing the release.
13:58:29 <gjanssens> Yay!
13:58:34 <gjanssens> ;)
14:04:47 *** Topcat has quit IRC
14:06:28 *** Topcat has joined #gnucash
14:08:12 *** benoitg has joined #gnucash
14:11:29 *** aqua___ has quit IRC
14:11:55 <gjanssens> make distcheck fails on my system :(
14:12:06 <gjanssens> I suspect this is due to guile 2
14:12:23 <gjanssens> I have created a bug to remind us to look into this:
14:12:26 <gjanssens> https://bugzilla.gnome.org/show_bug.cgi?id=709589
14:13:35 <john> Does make check pass? That could be a blocker, as folks on Ubuntu and a couple of other bleeding-edge distros don't have guile 1.8.
14:19:02 <gjanssens> make check is currently running, but I expect it to fail as well
14:19:57 <gjanssens> guile spews out a syntax_transformer error on N_
14:20:12 <gjanssens> Backtrace:
14:20:14 <gjanssens> In ice-9/boot-9.scm:
14:20:15 <gjanssens> 157: 8 [catch #t #<catch-closure 28c3ce0> ...]
14:20:17 <gjanssens> In unknown file:
14:20:18 <gjanssens> ?: 7 [apply-smob/1 #<catch-closure 28c3ce0>]
14:20:20 <gjanssens> In ice-9/boot-9.scm:
14:20:21 <gjanssens> 63: 6 [call-with-prompt prompt0 ...]
14:20:23 <gjanssens> In ice-9/eval.scm:
14:20:24 <gjanssens> 432: 5 [eval # #]
14:20:26 <gjanssens> In ice-9/boot-9.scm:
14:20:27 <gjanssens> 2320: 4 [save-module-excursion #<procedure 2911d00 at ice-9/boot-9.scm:3961:3 ()>]
14:20:29 <gjanssens> 3968: 3 [#<procedure 2911d00 at ice-9/boot-9.scm:3961:3 ()>]
14:20:30 <gjanssens> In unknown file:
14:20:32 <gjanssens> ?: 2 [load-compiled/vm "/home/janssege/.cache/guile/ccache/2.0-LE-8-2.0/kobaltnet/janssege/Development/Builds/gnucash-f19-trunk/gnucash-2.5.5/src/app-utils/test/test-load-module.go"]
14:20:33 <gjanssens> In /kobaltnet/janssege/Development/Builds/gnucash-f19-trunk/gnucash-2.5.5/_build/src/app-utils/test/../../../../src/app-utils/test/test-load-module:
14:20:35 <gjanssens> 14: 1 [#<procedure 2d5d6c0 ()>]
14:20:39 <gjanssens> In unknown file:
14:20:41 <gjanssens> ?: 0 [#<syntax-transformer N_> "foobar"]
14:20:43 <gjanssens> ERROR: In procedure #<syntax-transformer N_>:
14:20:45 <gjanssens> ERROR: Wrong type to apply: #<syntax-transformer N_>
14:20:47 <gjanssens> Which is absolutely like chinese to me
14:20:54 <gjanssens> I'll note that this is not new with the gsettings code
14:21:07 <gjanssens> I ran into it before, but forgot to report it back then
14:21:19 <gjanssens> So, yes, this is a blocker for 2.6
14:21:39 <gjanssens> Because guile 2 support is one of the major features of this release
14:21:42 *** Topcat has joined #gnucash
14:22:08 <john> Ah, not new. OK, today's release can keep going.
14:22:15 <gjanssens> Yep
14:23:14 * gjanssens has three machines at the same time working on gnucash right now
14:23:23 <gjanssens> my linux box is running make check
14:23:47 <gjanssens> my OS X laptop is ploughing through a jhbuild (webkit currently)
14:24:01 <gjanssens> my Windows VM is running a plain build...
14:24:07 <gjanssens> Poor fileserver...
14:24:42 <gjanssens> confirmed - make check fails as well
14:24:46 <gjanssens> Same error
14:27:44 *** Topcat has quit IRC
14:28:41 *** Topcat has joined #gnucash
14:38:20 *** Topcat has quit IRC
14:39:56 *** Topcat has joined #gnucash
14:53:57 *** Topcat has quit IRC
15:06:37 <john> Release is tagged, but the tag doesn't seem to have made it to Github.
15:10:10 *** jmd has quit IRC
15:10:55 *** johey has left #gnucash
15:11:12 *** shaun has quit IRC
15:13:57 <john> Wait, yes it has. It just took a while.
15:31:16 *** Krzysiek_K has left #gnucash
15:39:36 *** Krzysiek_K has joined #gnucash
15:42:24 <gjanssens> Bad news: build fails on Windows
15:42:57 <gjanssens> Apparently gsettings key names can't be longer that 32 characters
15:43:17 <gjanssens> I suspect a Windows registry restriction
15:43:44 <gjanssens> I'll go throught all parameters and fix those that violate this limitation
15:58:11 <john> I thought that you'd already tested on Windows.
15:58:59 <gjanssens> The error pops up while building
15:59:15 <gjanssens> The tests I did hadn't used long preference names
16:00:04 <gjanssens> I have a patch ready, there's only one preference violating the constraint
16:00:11 <mikee> Geert: Compiling...
16:00:20 <gjanssens> Rebuilding on windows now to verify other errors
16:05:45 <mikee> Geert: Compiling again after distclean...
16:08:14 *** Krzysiek_K has left #gnucash
16:25:07 <mikee> * 21:21:55 OTHER <GLib-GIO> Settings schema 'org.gnucash.dialogs.import.hbci' is not installed
16:25:11 <mikee> Trace/breakpoint trap (core dumped)
16:25:30 <mikee> My build options are:
16:25:40 <mikee> OPTIONS="--enable-debug $PREFIX --with-html-engine=webkit --enable-python --enable-ofx --disable-dbi"
16:26:41 <gjanssens> Rats, I ran into a similar issue with ofx
16:27:01 <gjanssens> The problem is that some settings schemas are only included conditionally
16:27:27 <gjanssens> The schema missing on your system is only installed when aqbanking is enabled
16:27:56 <gjanssens> But the migration script doesn't take this into account
16:28:07 <mikee> I havn't been testing with all features. Guilty.
16:28:40 <gjanssens> It will check for each and every convertible preference
16:29:21 <gjanssens> Regardless of what is actually enabled at configure time
16:30:29 <mikee> I'll try a vanilla confiure but I'll have to install the dbi stuff first.
16:30:52 <gjanssens> Probably that won't work either
16:31:00 <gjanssens> Aqbanking is not enabled by default
16:31:15 <gjanssens> But you must have had it enabled in the past or in a stable version
16:31:28 <mikee> Since I re-installed fedora I'm still slowly fighting various databse issues, involving GIS software.
16:31:40 <gjanssens> Otherwise you wouldn't have its preferences stored in gconf
16:31:50 <gjanssens> *pondering*
16:32:25 <gjanssens> I will have to modify the base file that defines the migratable preferences
16:32:38 <gjanssens> And have it preprocessed by configure
16:32:51 <gjanssens> based on what optional features are enabled
16:33:22 <gjanssens> So it will only attempt to migrate preferences for features that are actually enabled
16:33:51 <gjanssens> Question is: should this still be done for this release ?
16:34:45 <gjanssens> btw mikee, when did you get this error ?
16:34:55 <gjanssens> When you tried to run gnucash I suppose ?
16:34:58 <mikee> Running.
16:35:03 <mikee> Installed fine
16:35:48 <mikee> I'll add --enable-aqbanking and try again.
16:51:11 <mikee> Same result. Trace/breakpoint trap (core dumped)
16:51:37 *** aqua___ has joined #gnucash
16:51:46 <gjanssens> With the same missing schema error ? Or another one ?
16:52:45 <gjanssens> Preprocessing in Configure won't work :(
16:53:21 <gjanssens> That would comment out lines using # marks
16:53:31 <gjanssens> Which won't work in xml files
16:53:47 <mikee> Same.
16:54:08 <mikee> OTHER <GLib-GIO> Settings schema 'org.gnucash.dialogs.import.hbci' is not installed
16:55:00 <gjanssens> Can you check if the schema file exists in $installdir/share/glib-2.0 ?
16:57:05 <mikee> Not there.
16:57:42 <mikee> csv., generic qif only.
17:00:02 <mikee> --enable-hbci should be deprecated but I'll try it anyway.
17:05:11 <john> It's getting late. Should I back out the release and try again with r23209?
17:05:55 <mikee> Same result. <GLib-GIO> Settings schema 'org.gnucash.dialogs.import.hbci' is not installed
17:06:30 <mikee> and no hsbci file in gnucash-trunk/share/glib-2.0/schemas/
17:07:01 <mikee> john: I'm far too drunk to comment.
17:07:25 <gjanssens> mikee: which is very odd, I don't see how the schema installation can be skipped with --enable-aqbanking
17:07:43 <gjanssens> john: I have just pushed a commit that fixes the windows build
17:08:15 <gjanssens> the one fixing the pref name that's too long
17:08:59 <gjanssens> To get past the missing org.gnucash.dialogs.import.hbci schema, I propose to disable migration of the aqbanking preferences for now
17:09:19 <gjanssens> together with ofx, these are the only two optional features that use preferences
17:09:30 <gjanssens> That's a 5 minue fix
17:09:41 <john> OK.
17:09:45 <gjanssens> And we can look into more detail at the issues for 2.5.7
17:10:22 <mikee> Have I missed anything else to check?
17:13:29 <john> I guess we need to try again after Geert pushes his changes.
17:13:58 <gjanssens> you could check if our aqbanking import module is building for you
17:14:51 <gjanssens> in $builddir/src/import-export/aqbanking
17:15:15 <gjanssens> I have pushed the change I proposed
17:16:19 <gjanssens> With this change it shouldn't matter if aqbanking is enabled or not
17:16:40 <gjanssens> But I still would be interested to learn where it fails
17:16:50 <gjanssens> I can't reproduce here currently
17:19:07 <mikee> I see the source files but no object files.
17:19:26 <gjanssens> So for some reason aqbanking support is not building on your machine
17:20:09 <mikee> Seems so.
17:20:52 <mikee> No idea why though.
17:22:45 <gjanssens> Neither do I :(
17:23:35 <gjanssens> I'm looking at configure.ac, but it seems to me it should either fail due to missing dependencies or succeed
17:23:43 <mikee> I had --disable-dbi. I'll build it again without.
17:24:04 <gjanssens> If it succeeds, your makefiles should start building aqbanking support
17:24:04 <john> That should be skew to aqbanking.
17:24:42 <gjanssens> (my comment was a continuation of my previous sentence, not an ack of your --disable-dbi message)
17:25:07 <mikee> OK
17:25:58 <gjanssens> Anyway, try with the latest trunk
17:26:17 <gjanssens> Which should work regardless of aqbanking support
17:26:33 <gjanssens> If that works, John can retag the build
17:27:09 <gjanssens> We can look at the aqbanking (and ofx) issues later
17:27:46 *** john_ has joined #gnucash
17:28:37 <gjanssens> john, the gnucash-git build I started on my OS X laptop earlier this evening is still building ORBit
17:28:49 <gjanssens> (And fails at that)
17:29:14 <gjanssens> Should I rerun the setup ?
17:30:17 <john_> Failing is odd, but since I haven't pushed the changed moduleset yet, it's not surprising that it builds it. No, the setup doesn't have anything to do with it.
17:30:27 *** mikee_ has joined #gnucash
17:30:48 <mikee_> Need some swords.
17:31:53 *** aqua___ has quit IRC
17:32:09 <gjanssens> heh, I'm starting to see double
17:32:26 <gjanssens> there's mikee and mikee_ now, as well as john and john_
17:32:38 *** john has quit IRC
17:32:38 *** Jimraehl1 has quit IRC
17:32:38 *** mikee has quit IRC
17:32:38 *** john_ is now known as john
17:32:45 * gjanssens didn't drink though
17:32:48 <mikee_> Used to it now. Beer assisted.
17:32:55 *** mikee_ is now known as mikee
17:33:08 <mikee> @op
17:33:09 *** gncbot sets mode: +o mikee
17:33:11 <gjanssens> And now they're gone again :)
17:33:30 <john> OTOH, you can tell it to ignore the problem. You'll have to build Gnucash by hand from a jhbuild shell because the requirements won't be met, or you could start over with skip GConf2.
17:33:31 <mikee> Twice the work in half the time.
17:33:35 <john> @op
17:33:35 *** gncbot sets mode: +o john
17:34:22 <gjanssens> mikee :)
17:35:47 <mikee> Yay. running.
17:35:48 <gjanssens> john, I'll try some other time.
17:36:05 <mikee> I need a drink after that.
17:36:06 <gjanssens> I've seen enough builds for today
17:36:36 <gjanssens> mikee: good!
17:37:18 <john> Good indeed. I've already untagged, so I'll retag as soon as the make distcheck is done.
17:37:38 <gjanssens> Ok
17:37:58 <gjanssens> I'll leave the rest of the release to you guys
17:38:13 <gjanssens> It's late here, and I'm ready to go to bed
17:38:31 <gjanssens> We've made good progress
17:38:46 <gjanssens> And found some issues that need some extra attention
17:38:56 <gjanssens> Thanks for beta testing my work :)
17:39:07 <mikee> Mostly you. No problem.
17:39:19 <mikee> Cheers all. Sleep now.
17:39:25 <gjanssens> Bye
17:39:36 *** gjanssens is now known as gjanssens-afk
17:39:42 <john> G'night.
18:38:07 *** benoitg has quit IRC
19:32:04 *** fell__ has joined #gnucash
19:32:04 *** gncbot sets mode: +o fell__
19:40:07 *** fell_ has quit IRC
19:43:10 *** Krzysiek_K has joined #gnucash
20:02:47 *** Krzysiek_K has joined #gnucash
20:27:30 *** warlord has quit IRC
20:48:36 *** ErKa has quit IRC
20:51:49 *** Krzysiek_K has left #gnucash
21:06:38 *** ErKa has joined #gnucash
21:31:58 *** ErKa has quit IRC
21:46:18 *** warlord has joined #gnucash
21:46:19 *** gncbot sets mode: +o warlord
22:06:01 *** bpburns has joined #gnucash
22:14:00 *** kpreid has quit IRC
22:14:54 *** kpreid has joined #gnucash
23:30:58 *** fell__ has quit IRC
23:44:36 *** jmd has joined #gnucash
23:54:34 *** jmd has quit IRC