//
// 2) FileDescriptorWatchers don't have HUP events.
//
-// As a POSIX-specific hack we MSG_PEEK and generate a syntetic HUP event
+// As a POSIX-specific hack we MSG_PEEK and generate a synthetic HUP event
// if we got 0 bytes.
//
// 3) You can't cancel a PostTask'd or PostDelayedTask'd task.
// cancelled if only there was a mechanism for it. The fingerprint in simply
// a monotonically increasing integer which is stored in the timer and also
// associated with a pending timeout. Whenever the timer is reconfigured
-// (delay updated, or timer disabled) the fingerpring is updated. Timeout
+// (delay updated, or timer disabled) the fingerprint is updated. Timeout
// callbacks with mismatching fingerprints are ignored.
//
// Originally we simply scheduled a task from the I/O thread to the UI thread
// hairier than I'm comfortable with... I think it'd be a better idea to run
// the mainloop fully in the I/O thread, have the resource sets live in the I/O
// thread also and have proxy object for them attached to the media backend
-// obejcts which would send requests and receive notifications to/from the
-// resource sets. We could avoid most of the threading related compliations
+// objects which would send requests and receive notifications to/from the
+// resource sets. We could avoid most of the threading related compilations
// for pumping the mainloop...
//
// Additionally to the basic mainloop adaptation we also set up the Murphy
void MurphyMainloop::Unregister(void* data) {
MRP_UNUSED(data);
- mrp_debug("unrgistering mainloop with data %p", data);
+ mrp_debug("unregistering mainloop with data %p", data);
}
bool MurphyMainloop::setupMainloop() {