|3 months 16 hours||build: s/DISTCHECK_CONFIGURE_FLAGS/AM_DISTCHECK_CONFIGURE_FLAGS/
DISTCHECK_CONFIGURE_FLAGS is supposed to be used from command line, so
using the automake counterpart instead.
|3 months 23 days||build: improved win32/win64 installer
Took some advices from the ntdisp project for the windows installers:
included libgcc and the (old missing) bzip2 DLLs and modified
|3 months 23 days||doc: updated NEWS|
|3 months 23 days||build: bumped version to 0.7.4|
|3 months 23 days||doc: updated po files|
|3 months 23 days||build: updated gschemas.compiled|
|3 months 23 days||doc: improved README stuff
Merged README.xml and README.md, added pango-cairo dependency and
(hopefully) improved grammar.
|3 months 23 days||adg: "transfer none" on get_cairo_path()
Having "const" as return value does not automatically set it to
"transfer none", so setting it explicitely in the docblock.
This should solve issue #110: http://dev.entidi.com/p/adg/issues/110/
|3 months 23 days||build: gitignore ./compile|
|5 months 14 days||doc: refactored to be more mainstream
Dropped custom XSL and merged latest gtk-doc.make into
Makefile.am.gtk-doc: there is no more need to have XML 1.0 compiant
pages because the new online website is based on silverstripe-gtkdoc
(http://dev.entidi.com/p/silverstripe-gtkdoc/) that accepts HTML code.
Corrected minor bugs during the process:
* removed wikilinks in @short_description because they are not properly
handled by gtk-doc;
* removed trailing dots from @short_description for consistency;
* added index of new APIs introduced by 1.0 version;
* added index of deprecated APIs (actually empty);
* merged docs/gnu into docs/adg;
* removed docs/common.
|7 months 29 days||doc: updated Lua bindings link in README.xml|
|8 months 1 day||doc: corrected NEWS.xml typo|
|8 months 2 days||build: bumped version to 0.7.3|
|8 months 2 days||doc: updated NEWS.xml|
|8 months 2 days||build: depends on cairo-gobject if introspection is enabled
If introspection is used, GObject support in cairo must be enabled
because gobject-introspection package relies on the presence of this
wrapper library. That support is present since cairo 1.10.0.
|8 months 3 days||build: stripped --internal arg from glib-genmarshal
On CentOS 5 (a fairly old GNU/Linux dist), the --internal flag of
glib-genmarshal generates the following error while linking:
/usr/bin/ld: .libs/libadg_1_la-adg-container.o: relocation R_X86_64_PC32
against `adg_marshal_VOID__OBJECT' can not be used when making a shared
object; recompile with -fPIC
|8 months 3 days||adg: provided cairo-gobject fallbacks
Consider cairo-gobject entirely optional and implement fallbacks for
every GObject wrapper provided by it, i.e. only for cairo_pattern_t.
|8 months 3 days||build: make --internal of glib-genmarshal optional
glib-genmarshal --internal has been introduced in glib 2.13.0. Check the
GLib version and eventually strip this flag from the makefiles.
|8 months 3 days||adg: make cairo-gobject optional
cairo-gobject has been added to cairo only recently (2010), so including
it only optionally. In case it does not exist, use the yet existing
fallbacks (or provide the missing ones if there will be).
|8 months 3 days||build: downgraded intltool dependency to 0.35.0
My web server is based on CentOS 5 and has intltool 0.35.0 installed,
and ADG do not rely on a specific i18n feature.
|8 months 6 days||cpml: alias CpmlVector to CpmlPair in introspection data
Renaming the typedef from struct _CpmlPair to CpmlPair properly aliases
CpmlVector to CpmlPair in the gir file.
|8 months 8 days||adg: cosmetic refactoring on adg-canvas.c
Corrected GObject introspection annotation where clearly wrong.
|8 months 11 days||adg: changed comment starting with @
When a line inside a docblock start with @, GObject introspection and
gtk-doc try to consider it a parameter name (it looks for an ending :).
|8 months 11 days||adg: removed adg_matrix_new() API
That function will make GObject introspection check for the
unexistent AdgMatrix type.
|8 months 12 days||Updated damned gschemas.compiled|