[Pidgin] #15755: Adium crash as a result of bug in purple_xfer_set_local_filename

[Pidgin]

#15755: Adium crash as a result of bug in purple_xfer_set_local_filename
--------------------+------------------------
 Reporter:  fain    |       Owner:
     Type:  patch   |      Status:  pending
Milestone:          |   Component:  libpurple
  Version:  2.10.7  |  Resolution:
 Keywords:          |
--------------------+------------------------
Changes (by fain):

 * status:  pending => new


Comment:

 This setter pattern is actually very common in Objective-C (and probably
 other languages). Here's someone describing why it's a problem to free an
 object before copying it without checking if they're the same:
 http://stackoverflow.com/questions/12369787/correctly-override-setter-in-
 objective-c

 The API doesn't specify that the argument can't be the same as the
 existing value. We *could* update the header files to include that (IMHO
 needless) restriction, or, just as easily, we could fix the setter
 implementations. And in doing so, not force Adium and other clients to
 perform that check themselves whenever calling a setter.

 If it makes things easier for integration I could just create a patch for
 this entire file and attach it to the ticket; let me know.

-- 
Ticket URL: <https://developer.pidgin.im/ticket/15755#comment:4>
Pidgin <http://pidgin.im>
Pidgin