Re: [Sugar-devel] Sugar + OLPC mesh network selection logic
- From: David Farning <dfarning sugarlabs org>
- To: Chris Ball <cjb laptop org>
- Cc: bens alum mit edu, Gary C Martin <gary garycmartin com>, sugar-devel <sugar-devel lists sugarlabs org>, networkmanager-list <networkmanager-list gnome org>
- Subject: Re: [Sugar-devel] Sugar + OLPC mesh network selection logic
- Date: Tue, 23 Jun 2009 15:47:46 -0500
On Mon, Jun 22, 2009 at 10:08 PM, Chris Ball<cjb laptop org> wrote:
> Hi Gary,
> > Hmmm, how about when a class of kids first boot up and
> > _automatically_ connect to some random, unfiltered, strangers
> > open/public wifi next to the school, rather than following the
> > teachers instructions to attach to an encrypted school AP running
> > through a firewall?
> I'd give the same response: you say an authority figure might want
> something else to happen, but the authority figure hasn't preloaded
> the encrypted network as a favorite network on the laptop (much more
> sensible than expecting young children to type in WPA passwords), and
> so all we have to go on is what we think the will of our user is.
> In a conflict between what the user might want and what an imagined
> authority figure might want, I think the user should win.
I would strongly support the notion that we separate that notions of
how Sugar 'should' behave and how 'authority figures might want' Sugar
The issue is that every 'authority figure' has a different idea of how
they want Sugar to behave in their particular situation. This
basically involves customizing the code base for each deployment. The
overhead to maintain these customizations grows quickly.
Sugar Labs should focus on making a standard ' best' behavior. I
support the the Apple like check box. I would just like to make sure
we get to the decision for the right reasons.
] [Thread Prev