Re: Bug in autoconnect?

[Pat Suwalski <pat suwalski net>]

On Mar 18, 2010, at 20:49, Dan Williams wrote:
>> Is that the intended behaviour? To me, autoconnect indicates that it's 
> 
> Yes.
> 
>> to be automatically connected unless I tell it otherwise, which is what 
>> deactivateConnection() would logically do.
> 
> You're probably looking for the Disconnect() method on the Device
> object.  That will take the device down and prevent any other connection
> from being activated on that device until you (a) tell it to activate
> something, or (b) suspend/hibernate.  Maybe we'll clean this up for 0.9
> and make Disconnect/DeactivateConnection work the same way, but for 0.8
> we decided to preserve backwards compatible behavior.

I'll experiment with that, thanks for the pointer.

>> Is the correct way to change the autoconnect variable of the connection, 
>> emit Updated(), and then call DeactivateConnection()?

FWIW, I did implement this, and it works as well.

--Pat