Re: G-Wrap 1.9.8 released

[Andreas Rottmann]

Thomas Klausner <address@hidden> writes:

> On Sat, May 19, 2007 at 05:57:01PM +0200, Andreas Rottmann wrote:
>> Yes. You found a bug here, the automake conditional for GLib would
>> always be disabled, no matter what the results of the check. This
>> should be fixed in bzr[0] now - it would be great if you could test it
>> (and gnucash using it, if you have time).
>
> When running autogen.sh with
>   checking for autoconf >= 2.59 ... found 2.61, ok.
>   checking for automake >= 1.5 ... found 1.10, ok.
>   checking for libtoolize >= 1.4.0 ... found 1.5.22, ok.
> I get:
> + running automake -a -c...
> configure.ac:11: installing `./missing'
> configure.ac:11: installing `./install-sh'
> doc/Makefile.am:1: installing `doc/texinfo.tex'
> g-wrap/Makefile.am: installing `./depcomp'
> guile/examples/Makefile.am:17: wildcard gw-*: non-POSIX variable name
> guile/examples/Makefile.am:17: (probably a GNU make extension)
> guile/g-wrap/gw/Makefile.am:26: addprefix standard, .h .c .scm:
> non-POSIX variable name
> guile/g-wrap/gw/Makefile.am:26: (probably a GNU make extension)
[...]

This is are warnings from automake-1.10, earlier versions had these
checks disabled. autogen.sh now passed -Wno-portability to automake to
prevent the warnings.

> I also patched g-wrap/Makefile.am to precreate the /ws subdir so that
> installation works (Greg sent this as a Makefile.in patch, attached is
> a Makefile.am patch).
>
Applied, thanks.

> I used libffi from the 1.9.8 release. I translated the pkgsrc
> configure patch to a configure.ac patch, please include it in the next
> release (attached).
>
I've now updated libffi from GCC SVN and have hand-merged this, please
check if I've done this right.

There is a 1.9.9 release candidate at
http://stud3.tuwien.ac.at/~e9926584/tmp/g-wrap-1.9.9-RC1.tar.gz;
everybody please give it a try to see if the issues you reported are
indeed resolved.

> With the bzr version, some more files are installed:
>       lib/g-wrap/modules/libgw-guile-gw-glib.*
>       share/guile/site/g-wrap/gw-glib.scm
> and gnucash-2.0.5 gets a bit further:
>       Backtrace:
>       In current input:
>          1: 0* [gnc:report-menu-setup]
>          ?: 1  (letrec (# # # ...) (gnc:add-extension income-expense-menu) 
> ...)
>       In /usr/pkg/share/gnucash/guile-modules/gnucash/report/report-gnome.scm:
>        118: 2* [gnc:hook-run-danglers "hook_report" . #f]
>       In /usr/pkg/share/gnucash/scm/hooks.scm:
>         22: 3  [gnc:hook-run-danglers-real "hook_report" #f]
>
>       /usr/pkg/share/gnucash/scm/hooks.scm:22:3: In procedure 
> gnc:hook-run-danglers-real in expression (gnc:hook-run-danglers-real hook 
> args):
>       /usr/pkg/share/gnucash/scm/hooks.scm:22:3: Wrong type argument in 
> position 2: #f
>
Unfortunatly I've no idea where this comes from, as I'm not familiar
with the GnuCash sources. If anybody has an idea where this error
stems from, I'd be real happy ;-).

Regards, Rotty
-- 
Andreas Rottmann         | address@hidden      | address@hidden | address@hidden
http://rotty.uttx.net    | GnuPG Key: http://rotty.uttx.net/gpg.asc
Fingerprint              | C38A 39C5 16D7 B69F 33A3  6993 22C8 27F7 35A9 92E7
v2sw7MYChw5pr5OFma7u7Lw2m5g/l7Di6e6t5BSb7en6g3/5HZa2Xs6MSr1/2p7 hackerkey.com

Any technology not indistinguishable from magic is insufficiently advanced.
   -- Terry Pratchett