On 20 November 2012 02:04, Trever <trr...@gmail.com> wrote:

If it is, it must be very hidden from the general public.

--
Ottavio

Also the simple removeal of an external screw to allow flashing your own
firmware/keys.  Not aware of such an easy feature on other existing Chrome
devices.

2012/11/19 Trever <trr...@gmail.com>

I would say false -- all Chrome OS devices have been just as friendly for
hobbyists and tinkerers by allowing for developer mode out of the box, i.e.
no need for any special trickery to "jailbreak" a device and get root
access to it. We're only careful to make sure that have _both_ a
hobbyist-friendly device _and_ a secure-by-default device for the users who
just want a simple and secure Chrome OS system.

Not quite sure I follow what you mean here, the guts aren't exactly kept
secret. Maybe none of the teardown websites have done articles covering it,
since it's not quite as exotic hardware and a fairly normal x86 system on
the inside (with a few Chrome OS-specific features).

The only piece of the current x86 firmware that is not open source
(provided as a binary) is the memory initialization code (MRC) that comes
from Intel. The rest of coreboot/u-boot/vboot are all open and public and
can be found on git.chromium.org.

Note that the Atom-based x86 machines shipped with proprietary firmware
that is not open source. Only the Samsung 550 Chromebook and the Chromebox
(and the new $200 Acer) have the fully open coreboot firmware stack.

-Olof

What I mean is just that a person can unscrew an external screw to get RW
access to the RO firmware, and this is clearly documented in the case of
Snow.  Very cool.

How do I accomplish the same thing on, say, Stumpy?  I have a disassembled
Stumpy and haven't yet found the place to effect the firmware.  In fact,
didn't even know it was possible until recently.  See also:  
https://groups.google.com/a/chromium.org/forum/#!topic/chromium-os-di...

It would be more hobbyist friendly if this information were clearly
documented as it is with Snow, you know?

To be able to (eg.) write in one's own public key for a non-Google full
verified boot is a nice feature.  I agree about the "openness" in principle
of the machines and the codes.  To say that one Chrome model is more
hackable than another doesn't imply the other units are closed or
"unhackable" in the way some competitors might do.

I do think that in practical terms, the minimalist core/uboot makes hacking
a challenge, given present day documentation and fussiness of the
initialization code for the machines.  It makes *current* documentation
that much more important... and saying "just look at the source" really
isn't what some people want to do if they are (eg.) just trying to boot
other OS'es or want to sign their own distro using their own keys (all the
way through).

Again, not complaining about the openness in principle.  In fact, in
principle, the Chrome OS team has made some really cool stuff here.  That's
the point.  That's what makes getting this information to take advantage of
it all the more important, and all the more frustrating to not find it more
readily (not to mention that we are having to trouble you guys for this
info in mailing lists!  :-).

Thanks,

T

yes, if the general public is incapable of browsing a checked out
source tree or http://git.chromium.org/, then maybe they shouldn't be
hacking on it in the first place
-mike