[Tails-ux] Greeter: wording

Peter,

Since I had to look into many of the things you said, please take my
response lightly.

On 01/28/2015 20:25, Peter N. Glaskowsky wrote:
>> On Jan 28, 2015, at 5:11 PM, spencerone@??? wrote:
>> [ …]
>> I guess I just don't understand the technical requirements behind
>> installing TAILS using the installer, vs, dragging a ready to run OS
>> onto a portable storage media device. Presuming the OS can't be
>> portable like this [like an application], maybe what I am requesting
>> is having the installer be external.
>
> To summarize the obvious, drag-and-drop by itself can’t fully install
> an OS on a USB stick because the file-copy mechanism can’t do things
> like add boot blocks, change partition types, flag partitions as
> bootable, etc.
>
> It’s theoretically possible to have the file-copy mechanism recognize
>

Is this the same as Google's copy file mechanism, or is it just a
generic reference?

>
> that dragging a bootable ISO to a USB stick indicates the user’s
> desire to install the ISO on the stick. That could be used to shift
> control to something like the livecd-iso-to-disk application, which
> would work. That operation is both destructive and rare, however, so I
> wouldn’t support implementing it.
>

Fedora has both a destructive and a non-destructive method for using
their livecd-iso-to-disk tool, is this the same? If so, can files not
be destroyed, or do you mean something else by "destructive"? Also, is
rare bad?

>
> But you suggest something interesting:
>
>> You would drag the .iso onto your portable storage media device then,
>> after restarting your computer, point the boot dialog to the portable
>> storage media device to run the TAILS installer. If this is
>> technically possible, I don't see the need to involve host machine OS
>> file managers, right?
>
> This would be a different way to get a drag-and-drop experience for OS
> installation, as well as a way to boot multiple OSs from a single
> storage device… but I can see serious security issues if it’s that
> easy to get a new OS onto a storage device, and it implies a boot
> loader that is smart enough to dig into one or more ISO files at each
> reboot, or at least any new ones, and present them to users as
> potential candidates for booting.
>

Is this issue similar to why it isn't recommended to run TAILS on a
virtual machine, e.g., providing the host machine OS, or, in this case,
the other OSs on the portable storage media device, access to all of the
contents of any OS in question?

Also, presuming I understand enough about what you are saying, can a
protective shell be created around the OS that prevents the boot loader
from digging too deep while still allowing it access to the information
it needs to boot?

>
> So I don’t think I like this idea either.
>
> Still interesting, though. I’ll have to think about it some more.
> Thanks for the suggestion.
>
>
> .png
>

No worries, thanks for the consideration and the informative response :)

Wordlife,
SpencerOne