[Bugs] #1673 HIGH: discard network history does nothing

[Sugar Labs Bugs]

#1673: discard network history does nothing
-----------------------+----------------------------------------------------
    Reporter:  quozl   |          Owner:  tomeu            
        Type:  defect  |         Status:  new              
    Priority:  High    |      Milestone:  0.88.x           
   Component:  sugar   |        Version:  Git as of bugdate
    Severity:  Minor   |       Keywords:  r!               
Distribution:          |   Status_field:  New              
-----------------------+----------------------------------------------------

Comment(by quozl):

 None of you seem willing to show your suggestions as code, so I'm
 presuming you are looking for consensus.  I don't think consensus has been
 achieved.

 I'm happy with the commit in 0.84.15 on 2010-04-01 (fcb1cec), and have no
 great desire to fix it for 0.88 yet, but I'm sure we can get there
 eventually, once there is consensus on how it should be applied.  I shall
 wait for consensus.

 Regarding Tomeu's doubt that I will find in Sugar a loop with 3
 synchronous D-Bus calls like that one I'm introducing in my patch ...
 that's false, I'm copying this ''_disconnect_activate_cb'' code with
 slight modifications from ''_'''_'''deactivate_connection_cb'' in
 extensions/deviceicon/network.py ... and if you think that
 ''_'''_'''deactivate_connection_cb'' should be fixed then raise a separate
 ticket.  I don't think it should be fixed, because it's not broken.  Yes,
 mixing synchronous and asynchronous D-Bus calls is unwise, but it usually
 works, as can be seen in the existing code.  I propose a separate ticket
 for review and fix of use of synchronous D-Bus calls.

 Regarding a split point ... I agree that the split described by Tomeu is
 one option, but I don't see why it is the best option.  The unsplit patch
 was fully tested, and the 0.84 unsplit variant of the patch has had
 extensive testing in the OLPC XO-1 and XO-1.5 builds.  Also, I was
 especially asked by Tomeu in IRC to fix (and therefore test) the use of
 the dbus.PROPERTIES_IFACE constant.  Perhaps it should be dropped
 altogether if it causes a need to split.  I propose a separate ticket for
 that.

 Are there any further suggestions?  The new activity on this ticket is
 very encouraging.  I guess it is because the 0.88 deployments are
 affected?

-- 
Ticket URL: <http://bugs.sugarlabs.org/ticket/1673#comment:18>
Sugar Labs <http://sugarlabs.org/>
Sugar Labs bug tracking system