Remove get_object_by_unique_name again, and replace with a better-thought-out version...
commit3504ac8bc098b5872be747c151f304bbe80d2131
authorSimon McVittie <simon.mcvittie@collabora.co.uk>
Tue, 16 Jan 2007 17:04:37 +0000 (16 17:04 +0000)
committerSimon McVittie <simon.mcvittie@collabora.co.uk>
Tue, 16 Jan 2007 17:04:37 +0000 (16 17:04 +0000)
tree9eeb7e4dc80db9e8df89c5f585bd58ca67601ee8
parente8d766e1b2d5013e3f2c0a95d43b1dcb5eb00044
Remove get_object_by_unique_name again, and replace with a better-thought-out version of get_object

- Remove get_object_by_unique_name (introduced in 0.80rc1) - J5's right, the
  name is misleading
- Make get_object *not* follow name changes by default (a return to pre-0.80
  behaviour, and the same as 0.80rc1's get_object_by_unique_name())
- Add keyword argument follow_name_owner_changes (default is False): if True,
  it will follow name owner changes (like 0.80rc1's get_object() always did)
- Implement with modifications to ProxyObject so it resolves the unique name
  on construction, if required (activating it if necessary)
dbus/_dbus.py
dbus/proxies.py