[Kde-bindings] kdebindings-4.1.96, QtRuby/plasma-ruby-clock bootstrap

[Richard Dale]

On Friday 16 January 2009 08:02:36 Sebastian Sauer wrote:
> Rex Dieter wrote:
> > I'm working more on kdebindings-4.1.96 packaging for fedora, and have run
> > into what appears to be bootstrapping problem, wrt QtRuby and the
> > included ruby plasma applets.  cmake says:
> > rbuic4 not found, ruby-clock will not be created
> >
> > ...but rbuic4 is part of what is just being built.
> >
> > Shouldn't the kdebindings module be able to bootstrap itself? or am I
> > missing something and/or doing something wrong?
> >
> > -- Rex
>
> probably something like;
>
> Index: analog-clock/CMakeLists.txt
> ===================================================================
> --- analog-clock/CMakeLists.txt (revision 911762)
> +++ analog-clock/CMakeLists.txt (working copy)
> @@ -3,7 +3,7 @@
>  SET(UI_CLOCKCONFIG ${CMAKE_CURRENT_SOURCE_DIR}/clockConfig.ui)
>  SET(UI_CALENDAR ${CMAKE_CURRENT_SOURCE_DIR}/calendar.ui)
>  SET(UI_TIMEZONESCONFIG ${CMAKE_CURRENT_SOURCE_DIR}/timezonesConfig.ui)
> -FIND_PROGRAM(RBUIC4 rbuic4 PATHS ${BIN_INSTALL_DIR})
> +FIND_PROGRAM(RBUIC4 rbuic4 PATHS ${CMAKE_CURRENT_BINARY_DIR}
> ${BIN_INSTALL_DIR})
>  if(RBUIC4)
>         ADD_CUSTOM_COMMAND(OUTPUT
> ${CMAKE_CURRENT_BINARY_DIR}/analog_clock_config.rb COMMAND ${RBUIC4}
> ${UI_CLOCKCONFIG} -o ${CMAKE_CURRENT_BINARY_DIR}/analog_clock_config.rb
> COMMENT "Generating analog_clock_config.rb")
>         ADD_CUSTOM_TARGET(clockConfig ALL DEPENDS
> ${CMAKE_CURRENT_BINARY_DIR}/analog_clock_config.rb)
>
> would help?! But then it's still not guaranteed that rbuic4 got build
> already if "make -jMany" is used... maybe the attached patch would solve
> that (completly untested)?
Actually that clock example is obsolete, and won't even work until we can sort 
out the kde plugin versioning for ruby plugins. I'll remove it from the KDE 
4.2 branch and trunk.

The up to date example is in kdebase and i think for that i just checking the 
rbuic4 generated code into the svn. Maybe it would be better to use your 
solution above for that though.

-- Richard