Re: [Tails-ux] Wireframes for DAVE 2

Delete this message

Reply to this message
Autor: sajolida
Data:  
Dla: Tails user experience & user interface design
Temat: Re: [Tails-ux] Wireframes for DAVE 2
intrigeri:
> sajolida:
>> And you can see my (ongoing) HTML prototyping work on the production
>> website:
>
>> http://tails.boum.org/install/download_2


Thanks for reviewing this! I didn't dare asking you explicitly but I'm
very happy you do it :)

> s/loose/lose/>
> I'm not a big fan of "You might get hacked while using Tails if our
> servers have been compromised and are serving malicious downloads":
> it suggests that the verification step is worthwhile even if our
> website has been compromised, which is wrong. Perhaps replace
> "servers" with "download servers" or similar?


Yeah. I thought that in people's mind there's no difference between
"servers" and "download servers" and opted for the shorted version.

So what about "download mirrors"? Our audience might know what a mirror
is (cf. WikiLeaks back in the days, censorship events here and there,
etc.) plus the context should help understanding what this is about to
those who don't know yet.

> I'm not a big fan of "You might get hacked while using Tails if your
> download is modified by an attacker in your country or on your local
> network" + the link to the DigiNotar compromise either:
>
>  - The DigiNotar link suggests that the weakness verification protects
>    against is only about HTTPS, which is only the case when JS is
>    enabled (#12833). Now, when JS is disabled perhaps the "Why?" popup
>    can't be displayed anyway; if that's the case then you can ignore
>    this comment :)


Yeah, people with no JS get no popup :)

>  - This text seems to only address targeted attacks in a specific
>    country or against a specific user, but an adversary who can break
>    HTTPS can exploit it anywhere between the download servers and the
>    client. And when downloading using Tor, an adversary who can break
>    HTTPS can also exploit it close to the exit node being used.
>    We would detect such a compromise ourselves only when facing an
>    not-too-sophisticated adversary. I kinda remember having had this
>    discussion already, sorry if I was arguing in the other direction
>    last time ;)


As you can guess I don't remember this discussion at all.

I changed the sentence to:

You might get hacked while using Tails if your download is modified
on-the-fly by an attacker on the network.

So the attacker can be anywhere.

Better?