[Mageia-discuss] FSF anf UEFI SecureBoot

[Claire Robinson]

On 31/12/12 14:53, Ludovic V Meyer wrote:
> 2012/12/30 AL13N <alien at rmail.be>
> 
>> Op zondag 30 december 2012 21:17:38 schreef Ludovic V Meyer:
>>> Except it does let 3rd parties OS boot, at least on X86, since the norm
>>> mandate it.
>>> And for arm tablet, no one reacted when Apple, Acer, Samsung, Archos and
>>> lots of others locked down their devices, so trying to argue that we now
>>> expect them to be open would not work.
>>
>> actually, they didn't. you can root each of those iinm.
>>
> 
> Using 3rd exploit is not really what I call open, they are not supported,
> likely against DMCA most of the time, and IMHO not reliable.
> Not to mention that it requires a manual intervention on each device. If we
> take the example of Apple, they closed every hole after a while when it was
> practical to do,and used the existing leagal way to prevent them ( see in
> 2009,
> the update of the developper agreement ). And since I know you will surely
> talk of if, the DCMA ruling for jailbreaking is just for phone, because
> unlike France, telcos in USA do not have to unlock your phone after a few
> months.
> 
> Not to mention that afaik, despites them being "not closed" by your
> definition, stuff like Iphonelinux are all dead in the water.
> Cyanogenmod only exist because from time to time, Google do a code drop,
> and they still suffer from needing a custom fork of the kernel.
> 
> So if the goal is "to be able to run what I want on my device", that's
> something that can already be done for applications. What people should say
> is "running what I want provided no money directly leave my pocket, but I
> do not mind spending days figuring how to do it, cause I prefer spend 1
> week than giving 100 bucks".
> 
> this is about having a secure key hardcoded "burned" in the device, which is
>> both stupid and annoying. because since apps need to be secured too, too
>> many
>> people have access to the root key. which means the chance of leak is
>> higher.
>> which means that your devices need to be thrown out when the rootkey is
>> compromised or when it's deemed obsolete and a new key will be in place.
>>
> 
> The key is handled by Verisign, and since that's their jobs since around 18
> years, I think they are qualified to do it.
> How many time in 18 years was the root cert of Verisign be compromised ?
> 
> Also, you are totally wrong about throwing the device if the key is leaked.
> This happened to the PS3 due to the world-record breaking ignorance of Sony
> ( or one sub contractor ), and AFAIK, the PS3 all around the world still
> work ( and also, no one formally complained about gaming consoles being
> closed, despite some of them just being powerful PCs ). The same goes for
> various phones/tablet who have been broken this way ( like the Asus
> transformer, AFAIK ).
> 
> Burning a key in silicium is what Apple have been doing since a long time.
> That's also the modus operandi of TPM modules. They are used by several
> banking institutions as a way to make sure the harddrive is protected with
> bitlocker ( cause you do not want your highest executive laptops to be
> stolen and that this cause privacy and security issues ). IE, that is
> viewed as sufficient for FIPS certification and usage for military grade or
> banking grade security. And I am pretty sure the private key is stored in
> some HSM like the nShield solo or similar device.
> 
> Not everybody work like your client ( the one we talked about yesterday on
> IRC, if I am not wrong ). Some people take security seriously, and check
> what happens. But that's not security of the root key that matter, since no
> one ever asked for public scrutiny or a independent audit.
> 
> the thing here is that since you buy a device, it's yours and you can do
>> what
>> you want with it. why would you give other parties control over your
>> device?
>> it's stupid. there needs to be a way as an owner to decide which root keys
>> you
>> trust or not.
>>
> 
> You do not give control to another party, you delegate trust handling to
> another party.
> That's exactly what you do with a browser. Or your bank, or anything in
> life.
> 
> Again, the norm mandate to be able to disable secureboot on x86 and to
> choose the key. The whole petition is about those that do not follow the
> norm, and for those, the incentive was to not being Windows 8 certified. So
> as annoying this will be, that's the best way to find something that let
> you run Linux.
> 
> 
>>> And regarding using consumer protection channels, no one did anything to
>>> make anything move since one year despite being widely publicized on
>>> various blogs, so how is your proposal different ?
>>>
>>> Talk is cheap, if every people who proposed that ( for example, on
>> slashdot
>>> or various foras where nerds are discussing ), someone would have started
>>> the work by the time. No one did, and that's because everybody that would
>>> be serious enough know this is built on wrong assumptions.
>>
>> in the end talk is cheap and noone does anything about it. or rather
>> instead
>> of working together, all the companies who back the major linuxes decide
>> to go
>> down the easy route. (like subscribing into the microsoft program and using
>> their root key...)
>>
> 
> All plans that requires someone else to do anything is just a way to blame
> failure to someone else. If you delegate all your action to someone else,
> you lose the right to complain about this group not doing what you want.
> Only delusional fools would believe otherwise.
> 
> In fact, hardware not working on Linux is a decades old problem. We all
> have seen how boycott worked so well to have more hardware supported on
> linux, and how people happily trade freedom for convenience ( like nvidia
> drivers, printers, etc, etc ). People should just do a reality check from
> time to time before proposing the same plan again and again. Last time I
> checked, humans didn't evolve from goldfish, so maybe we could stop acting
> like them.
> 


Apathy is not the answer.

Claire