Re: [Barry-devel] Barry & PlayBook

[Chris F. <cd...@fo...>]

On Sun, Feb 05, 2012 at 07:21:06PM +0100, Nicolas wrote:
> You can find a barry update into my git repository :
> 
> http://repo.or.cz/w/barry/progweb.git
> 130cd47bf9f2ac7ce7fd91dbba6b9cccf5d39b0e
> 
> I have merged the bplaybook into the barry head.

Thanks... I found the update in your master branch.


> If I remember, you prefer that I write a new Probe Object specific for
> playbook device. But in fact, I think, that it's better if we have a
> commun Probe Object, then a ProbeBlackBerry and a ProbePlayBook object.
> So, ProbeBlackBerry and ProbePlayBook have the same hierarchy.

I still think that it belongs outside the Probe class.  The purpose
of the Probe class is to discover all available BlackBerry devices, and
fill a vector of ProbeResult objects to identify them.

But ProbePlayBook actually changes the state of the device it finds, and
then doesn't add anything to the ProbeResult list.

I don't mind if this functionality exists in the Barry library itself.
Perhaps Probe should add a ProbeResult for the PlayBook by default,
with a flag, to show that it has no database access.  Then you could
pass this ProbeResult into a PlayBookToNetwork() function, which simply
calls the two ControlMsg()'s to do the reset.

I suspect that someday soon, the PlayBook *will* have some kind of
database access, so Probe adding it to ProbeResult makes sense.  But
changing modes is outside Probe's scope in my opinion.

Wouldn't this make the code simpler?

- Chris