[p2p-research] Google Sniffing WiFi - OpenSource P2P Solutions

[M. Fioretti]

On Sun, May 16, 2010 17:15:14 PM +0200, Alex Rollin (alex.rollin at gmail.com) wrote:
> Perhaps some of you saw the Google blog post or heard about it through another
> outlet:
> 
> http://googleblog.blogspot.com/2010/05/wifi-data-collection-update.html
> 
> In the post Google says they did collect some "payload" data from
> Wifi networks as they drove their GMap Street View cars around...  I
> wonder if this would have happened if they would have simply
> released the code as open-source.  Then the world would have been
> looking it over since 2006 and the issue would have come up quite a
> bit sooner.

Of course it could have happened. Why not? What matters is what code
they *actually* ran on *their* cars. If they had released all the code
with a GPL license, somebody would have noted the bug and publicly
told Google, maybe also providing a patch, sure. Then what? They could
have very well said gee, thanks for telling us and thanks for the code
fix, and then continued to run the "buggy" version, if that's what
they wanted to do.

> It is the fact that it was done without any peer oversight that
> bothers me.

the *relevant* thing done without peer oversight here is driving the
cars around, not writing or debugging the software. It would have made
more of a difference if each car had also hosted an independent observer.

It's like e-voting: it's pointless to say "e-voting software can only
be open source" because what matters is only what code _actually_ runs
in each voting booth. And the only way to make sure that that code is
clean is to open the hardware and analyze the code there and then.

Marco