[XSCE] Re: Large groups of XO-1 do not work with access points

[James Cameron]

On Sat, Feb 08, 2014 at 10:20:29AM -0500, Tim Moody wrote:
> Does it matter if the xo has previously connected to the ap and
> stored connection info?

No, it doesn't matter.

There's a lot of misunderstanding.  I feel people aren't taking the
time to understand.

Restating the problem; with a stored connection, on reboot Sugar will
often not connect to it, because the network is not shown in network
neighbourhood, in turn because the scan results did not include the
network, in turn because the scan results were lost or corrupted, in
turn because of a flaw in the kernel or wireless firmware, in turn
triggered by the presence of mesh nodes.

> Could we avoid the scan by pre-populating the connection info?

No, Sugar will not initiate a connection to a wireless network name
unless the name is seen in the scan results.

> (I think you did this in tinycore.  This would also eliminate the
> variable of relying on NN to show us the AP.)

No, Tiny Core Linux is affected by the same problem to a lesser
degree, but as the user base for Tiny Core Linux is small to
non-existent, I don't care to worry about it.

> Since we are regularly reflashing, could we turn off mesh and
> prepopulate a connection to a standard SSID; in Haiti we used
> 'lekol'?

Turn off mesh; yes.  Prepopulate; no.

But I do not recommend turning off mesh.  Mesh is actually very useful
to the learners.

> If this strategy increases the likelihood of connection would it be
> possible to get fancier and have our boot script check for a
> connection to the known ap and if not found turn mesh back on?

Yes, this might help, but I think you should wait until we move from
trying to understand the problem and into selecting fixes and
workarounds.

It is taking a long time because we are doing it together, Terry and
I, and there is other work at OLPC that is also important.

I don't want you all to go off implementing quick workarounds that you
have to pull out.  ;-)  Especially if that prevents you from using and
testing a final solution.

-- 
James Cameron
http://quozl.linux.org.au/