[nzlug] Open Source Census

[Simon Bridge]

On Tue, 2008-04-29 at 09:31 +1200, Greg Fawcett wrote:
> I also baulked at the idea of letting an untrusted application run on my PC,
> reporting back who knows what. I'd have happily ticked boxes on a survey
> form.
This is the part of it that crunches -
When we download other code, the author pretty clearly states what we
get out of the effort of compiling and running. The more we have to do,
the more motivated we need to do it.

Some effort is usually made to *authenticate* the program to us, in the
way it is presented. Usually, we expect the distributor to pass, at
least, a basic snakeoil test.

So, I'll download and compile the madwifi drivers, though I'd rather
not, to get that blasted wifi going. The alternative is buying a
supported wireless nic. I know clearly what I'll got out of it and why
it is needed.

I install updates off ubuntu mirrors since they come signed and I get to
review what they are supposed to be for. I certainly would not install
an update for something I know I didn't install.

The survey program fails the snakeoil test. Consider:

Why do I need to install software to answer questions in a survey? What
are the questioners getting from this software that they cannot ask me
to my face? Why do I have to *hunt down* the justification?

Maybe it *won't* do any actual harm - it's still smart to avoid it.