The problem is not the concept. It is the implentation.
-
The requirement for using only one certificate authority (can you say
“monopoly” or “anti-trust”?) has been discussed as a major problem for
some.
-
The fact that an independent developer cannot get a certificate is
another. It could put those developers out of business. (And some of
them are very skilled driver developers - well-known and well-respected
in the community)
-
The fact that there is no automated way to get around this policy so
that automated tests will work is a MAJOR problem. My company is not
prepared to pay a lab monkey to sit and hit the F8 key every time the
system reboots as part of an overnight automated test in order for the
driver under test to be loaded.
-
Why target 64-bit vista with this policy? Driver availability is
already a problem that is getting in the way of widespread adoption of
the 64-bit platform. Why make that process any harder?
-
What about the learning process - people offering driver-development
classes with hands-on labs? Will those training companies need to allow
their students to use their PIC to sign the drivers they develop in
class? Or how about people who simply want to learn Windows driver
development as a marketable skill?
-
Shouldn’t the administrator of a system have ultimate control over
whether they want to allow an unsigned driver onto the system? With UAP
in vista? Sometimes a small in-house driver is required for the purposes
of testing other things (hardware diagnostics comes to mind). Why should
that type of driver need a signature?
-
As someone else mentioned there are small vertical markets that will
be greatly impacted by this as well. Customers in this market space
don’t require a signed driver. They can only get the driver from one
place so they know who it came from. Perhaps the concept of “small
vertical market” makes you think “not much money will be lost there if
they switch to a different platform”. Have you thought about the sum
total of all such markets? I bet that would be significant.
Beverly
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Henry Gabryjelski
Sent: Wednesday, January 25, 2006 12:13 PM
To: Windows System Software Devs Interest List
Subject: RE:[ntdev] X64 Windows Vista to require signed drivers
(Disclaimer: I have no internal knowledge of this, and all opinions
below are strictly my own and not necessarily Microsoft’s.)
Has anyone considered that malware, spyware, and rootkits are
increasingly loading in the kernel and becoming harder to detect? No
more kernel-based rootkits that aren’t trackable back to a corporation?
It just seems to me that having all kernel-mode bits signed seems like
it could greatly reduce this attack vector.
I’m just surprised this hasn’t come up already on the discussions.
.
-----Original Message-----
From: Roddy, Mark [mailto:xxxxx@stratus.com]
Sent: Tuesday, January 24, 2006 7:46 AM
Subject: RE: X64 Windows Vista to require signed drivers
Microsoft, along with everybody else in the consumer computer industry,
is focused on replacing that thing on top of your tv with something that
they build/manage that streams content into your home that is licensed
and generates revenue. Lots of revenue. From this discussion it is clear
that the downside of this is that other uses of their OS, from general
purpose server systems to various forms of low box count specialized
systems are going to lose out when decisions are made and there are
conflicting goals.
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Brown, Beverly
Sent: Tuesday, January 24, 2006 9:52 AM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] X64 Windows Vista to require signed drivers
This paragraph from the document sure does seem to support that:
“* Drivers must be signed for devices that stream protected
content. This includes audio drivers that use Protected User Mode Audio
(PUMA) and Protected Audio Path (PAP), and video device drivers that
handle protected video path-output protection management (PVP-OPM)
commands.”
As soon as I read that, I thought “So that’s the reason they’re doing
this.” I hate DRM. It protects a few media publishers and side effects
of things that get put in place to support DRM cripple everybody else -
ultimatley giving the consumer fewer choices which will lead to less
quality in the long run.
It seems to me that they could make this a requirement for only those
devices, though. Leave the rest of us alone.
Beverly
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Tony Mason
Sent: Tuesday, January 24, 2006 9:13 AM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] X64 Windows Vista to require signed drivers
Someone pointed out to me yesterday (in an offline conversation) that
the real issue here is NOT related to drivers, but rather to DRM.
Microsoft has to lock down the set of certificates in order to implement
their strong DRM policy (otherwise, you could add your own certs and use
them to bypass the DRM apparently.) I’m not certain if that’s correct,
but the person who said this to me is reliable - and it makes a certain
type of sense. It explains why they won’t use just any root cert (which
certainly doesn’t matter for drivers, but doe matter for DRM).
I’m sure the folks at Microsoft did speak to their customers to
determine the impact this would have. After all, it is difficult to
imagine that one could make such a fundamental decision like this
without consulting with key customers (imagine the sheer embarrassment
factor if you need to recant after taking a strong policy position such
as this one.) So, while it will be inconvenient for us, they have
apparently determined that this is an acceptable cost (and if you don’t
like it, please refer to “Figure 1” 
)
Regards,
Tony
Tony Mason
Consulting Partner
OSR Open Systems Resources, Inc.
http://www.osr.com
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Peter G. Viscarola
Sent: Monday, January 23, 2006 2:29 PM
To: ntdev redirect
Subject: RE:[ntdev] X64 Windows Vista to require signed drivers
I think the point about special-purpose drivers that are used in-house
or by third party companies in very specific markets is a good one.
There are TONS of these drivers, and requiring them to be signed is
nothing but a DISincentive for people to move to 64-bit Windows.
Sigh… I’m glad Microsoft is thinking about issues of driver security
and reliability, but I really wish they would enter into a dialog with
the community about these policies before mandating them. I dare say
that even THEY can’t think of every consequence of every proposed
policy.
Peter
OSR
Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256
You are currently subscribed to ntdev as: xxxxx@osr.com To unsubscribe
send a blank email to xxxxx@lists.osr.com
Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256
You are currently subscribed to ntdev as: unknown lmsubst tag argument:
‘’
To unsubscribe send a blank email to xxxxx@lists.osr.com
Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256
You are currently subscribed to ntdev as: unknown lmsubst tag argument:
‘’
To unsubscribe send a blank email to xxxxx@lists.osr.com
Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256
You are currently subscribed to ntdev as: unknown lmsubst tag argument:
‘’
To unsubscribe send a blank email to xxxxx@lists.osr.com
well, maybe Brasil too. China too.