On Tuesday 06 Aug 2013 17:00:26 Robert Hailey wrote:
> 
> On 2013/08/06 (Aug), at 5:24 AM, Matthew Toseland wrote:
> 
> > On Tuesday 06 Aug 2013 00:01:43 Robert Hailey wrote:
> >> How about avoiding the scary words, default to true (they already d/l it 
> >> from our server, right?) thusly:
> >> 
> >> [x] Report installation failures over HTTPS
> > 
> > No way.
> > 
> > Explicit, prior, informed consent.
> 
> I think it could be worded better, but isn't that already explicit (as 
> opposed to the act on installing implying they want it to work), prior 
> (before install), informed consent?
> 
> Consider this wording:
> 
> "[x] Send a report if installation fails (using HTTPS), so we can fix the 
> problem."
> 
> Surly if they are conscious enough to use Freenet and in such a sensitive 
> environment that they cannot risk the *potential* outgoing https connection, 
> then they will know to uncheck it.
> 
> I think it'd be more patronizing to presume they *don't* know what this means.

Nope. It's not safe to assume that everyone who has good reasons to use Freenet 
has a deep understanding of Freenet, or of networking, or of computers. And 
anything that potentially causes a severe privacy breach should be off by 
default. For the same reason that we still don't turn on opennet by default: 
it's something you have to ask the user about.

And anyway we don't need to ask all users - we can quite happily only ask say 
10% at random.
> 
> So too, I find it unlikely that one would execute freenet (or it's installer) 
> if they do not sufficiently trust the Freenet developers that the report 
> would not contain personal/identifiable information (as your wording lists 
> the contents).

No, your IP address contains identifiable information.
> 
> Are you concerned that someone could (for a particular network) manufacture 
> an installation failure (downloading java?) as a way to detect attempts at 
> installing Freenet?

Not necessarily. It's a fair assumption a significant number of users fail to 
complete it. We need to know when they fail to complete it - it could be 
anything from "I don't want to install Java" to "I didn't finish the wizard". I 
don't know if we need to record if they succeed, but there are enough failures 
here that Eve can happily just wait for connections.
> 
> >>> If you click yes, the Freenet installer will tell us whether your install 
> >>> of Freenet succeeded
> 
> > We only need feedback for a representative sample.
> 
> To me, it is far more unsavory, dangerous, and uninteresting to collect 
> positive samples (e.g. "success"!) than to automatically report a bug by 
> default.
> 
> > People do read. Most people have a fairly advanced reading age.
> 
> Fine, but why waste their time with a paragraph (or modal dialog) when it 
> only takes a sentence to accurately convey the idea?
> 
> Why optimize for the case that their life with end if we signal that they 
> can't setup the software, when we can optimize for our present reality?

Whatever. There's no point arguing with you.

For as long as I am around - which is a mere 8 weeks and 6 days from now - I 
will never, EVER add an unconfirmed or default-on callback option, nor will I 
allow anyone else to do so.

I believe I have the support of 99.9% of the users in this position.

Attachment: signature.asc
Description: This is a digitally signed message part.

_______________________________________________
Devl mailing list
[email protected]
https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl

Reply via email to