this looks like potentially causes a new cycle

Package python-pyparsing appears in cycle python-pyparsing/python3/bluez/gobject-introspection/python-pycparser
Package python3 appears in cycle python-pyparsing/python3/bluez/gobject-introspection/python-pycparser
Package bluez appears in cycle python-pyparsing/python3/bluez/gobject-introspection/python-pycparser
Package gobject-introspection appears in cycle python-pyparsing/python3/bluez/gobject-introspection/python-pycparser
Package python-pycparser appears in cycle python-pyparsing/python3/bluez/gobject-introspection/python-pycparser
Package python3 appears in cycle python3/bluez/gobject-introspection/python-pycparser
Package bluez appears in cycle python3/bluez/gobject-introspection/python-pycparser
Package gobject-introspection appears in cycle python3/bluez/gobject-introspection/python-pycparser
Package python-pycparser appears in cycle python3/bluez/gobject-introspection/python-pycparser

Stefan Brüns (StefanBruens) - over 4 years ago

bluez has BuildRequires: pkgconfig(libical)

at least the glib-doc should be build in a separate _mutlibuild, preferably the whole glib binding.

Dominique Leuenberger (dimstar) - over 4 years ago

reviewer

Just for reference, this is the cycle that needs to be broken off:

https://paste.opensuse.org/29722336

Easiest will probably be to build the -glib stuff as part of _multibuild

recheck

  8+%if "%{flavor}" == "glib"
  9+%define name_ext -glib
 10+%bcond_without glib
 11+%else
 12+%define name_ext %{nil}
 13+%bcond_with glib
 14+%endif

technically seen, bcond is the wrong usage here - the condition is not based on configuration wishes from the user, but purely based on the flavor picked (won't decline on it though, we have the same construct in other packages already too)