aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rw-r--r--libindicate/indicator.h31
1 files changed, 31 insertions, 0 deletions
diff --git a/libindicate/indicator.h b/libindicate/indicator.h
index 77ffb7d..3e2a803 100644
--- a/libindicate/indicator.h
+++ b/libindicate/indicator.h
@@ -133,6 +133,37 @@ void indicate_indicator_set_property_time (IndicateIndicator * indicator, const
const gchar * indicate_indicator_get_property (IndicateIndicator * indicator, const gchar * key);
GPtrArray * indicate_indicator_list_properties (IndicateIndicator * indicator);
+/**
+ SECTION:indicator
+ @short_description: A representation of state for applications
+ @stability: Unstable
+ @include: libindicate/indicator.h
+
+ An indicator is designed to represent a single instance of something
+ in your application. So this might be an IM or Email using #IndicateIndicatorMessage
+ or any other thing that is a small unit of information to pass on
+ to the user.
+
+ Indicators make no promises about how they are preceived by the
+ user, it's up to the listener to represent them in an intutive and
+ visually appealling way. But, what we can do is provide information
+ on the indicator to provide enough information for the listener
+ to do that.
+
+ Mostly this is done through properties. The only property that is
+ defined for the base indicator is the 'type' property. And this
+ is only available to set by creating a subclass of the
+ #IndicateIndicator object. It is assumed that you can look at the
+ definitions of the various subtypes to determine which properties
+ they support.
+
+ It may be that some users don't want to create objects for every
+ indicator as it could be a lot of overhead if there are large numbers
+ and there is already a data structure representing them all. In that
+ case it is recommended that you ignore the #IndicateIndicator object
+ tree in general and move to subclassing #IndicateServer directly.
+*/
+
G_END_DECLS
#endif /* INDICATE_INDICATOR_H_INCLUDED__ */