<sn0wmonster>
jackhill, in short, to my understanding, the only part of the phone the Librem5 can control is the software
<luke-jr>
sn0wmonster: huh? Neo900 doesn't have an open modem either
jcarpenter2 has quit [Read error: Connection reset by peer]
jcarpenter2 has joined #neo900
Hodges has joined #neo900
ArturShaik has quit [Ping timeout: 240 seconds]
jonwil has joined #neo900
ArturShaik has joined #neo900
<Ke>
sn0wmonster: I can't even parse, what you wrote
<Ke>
jackhill: right now there is a lot more info on Neo900 than on librem-5, but at least the form factor is completely different and librem-5 is expected to have a lot more computing power
<Hodges>
Personally, I'm torn on how to deal with the hype surrounding the Librem-5.
<Ke>
just don't let it get to you
<Ke>
but honestly speaking librem-5 is almost what I have been expecting someone to do for a very long time
<Ke>
I believe the goal is plausible and features sufficient
freemangordon_ has joined #neo900
<Ke>
they may do some things wrong initially and be less transparent than competitors though
<Hodges>
I suppose I just have a crisis of conscious because on one-hand I want it to succeed, but on the other it just seems like more counter-productive compromise that'll just deter better efforts.
<Ke>
there is indeed very small and fragmented niche that can't fit a lot of competition
<Ke>
but it's also about getting things done
<Hodges>
I just have the irreconcilable feeling that fully effort put into open hardware or open anything is being actively undermined. Definitely so in the US as far as I can make it.
<Ke>
I guess the magic should be in improving small scale manufacturing
<Hodges>
I can only imagine it working in a larger metropolitan city where public wifi has great coverage and community run cell towers are put up and managed/monetized similar to how things are handled with BTC or other blockchain based technologies.
<Hodges>
That's where I see the most difficulty.
<Hodges>
Any quality component with the right foot print, right interfaces, right features, etc# is locked away for exclusive use in commercially sold devices
<Hodges>
I can't see a viable product(form factor, luxury, convenience, price, etc#) for an end-consumer existing without that whole mnf paradigm changing or the free community hacking existing products in a more aggressive way.
Konsieur has joined #neo900
<Hodges>
One slightly related example. The earlier iPad displays. Absolutely amazing. Incredibly cheap by themselves, and all it took was some hackery and a simple break-out board to hook it up to anything via displayport.
<Hodges>
I feel like a thousand phones need broken apart and to find just one with a decent mostly-independent component ripped to pieces and repurposed for something else.
<Ke>
Hodges: someone taking it apart costs effort
<Hodges>
3.5-7in IPS, 1080P displays with a quality touchscreen and display connectivity like eDP should be ubiquitous on adafruit and the like.
<Hodges>
My background with this stuff is as having worked at a ITAD company that moved into the ERP/CRM space with a focus on embedded design and it's SaaS side playing along with it hand-in-hand.
<Hodges>
There is more to steal out of yesteryears garbage than people would imagine. Enough to create better niche devices than currently exist now, easily, economically, and sustainably.
<Hodges>
If a viable model were to be found for sourcing said components I know there to be a strong following for this sort of device - especially framed as being 'sustainable' or 'eco-friendly'.
<Hodges>
Because people don't know what to do with their freedom.
<Hodges>
Should that approach become a run-away success, what then when Apple, HTC, etc# learns that tens of thousands of last years phones that were thrown away across the entire US
<Hodges>
were funneled into a few select asset disposition firms
<Hodges>
Who pushed all of the necessary devices off for hacking would take their former products being re-sold commercially - or at all for that matter.
<Hodges>
It seems like a grey area
<Hodges>
But out of necessity, many communities exist around this. Retro video games being a big one.
chomwitt has joined #neo900
Hodges has quit [Ping timeout: 260 seconds]
Hodges has joined #neo900
<Hodges>
The reason I'm so heated about this is because I want to help and I believe I can but I don't have the right questions to ask.
<Hodges>
I'm meeting with my mentor later this week and I need some solid questions to take away some solid advice.
<Hodges>
Because he has the intelligence, connection, and capacity to really help push this forward with a wave of his hand if he believes in it and I believe in it.
<Hodges>
He is more than any man and can make the impossible happen. If I asked for 1000 N900s he can make it happen - however long, expensive, or cumbersome that might be for me manage.
<Ke>
I believe the most important first step is help liberating the SoC drivers and bootloaders
<Ke>
you also have to consider, what people want to get most of the niche crowd on board
<Hodges>
What's the motivating force, philosophically behind this? And why does every normie out there with their iPhone unknowingly share our same values? How do we push that on them?
<Hodges>
Because I think this needs adoption and more than anything traction
<Hodges>
and not bullshit developer hand-waiving to throw a few low-tier devs behind some random vaporware project for publicity on Phoronix.
<Ke>
if you want to sell soemthing, publicity is vital
<Ke>
puri.sm did that very well
<Hodges>
That's something else my role-model of a mentor has and can do.
<Hodges>
But we can't be wrong
<Hodges>
He has run several successful kickstarters
<Hodges>
In my time working with him we grew from 5 employees to 70 in 2.5yrs
<Hodges>
Successful exit at a high-caliber incubator with him holding all the equity and not giving a dime to anyone trying to steal his passion away.
<Hodges>
He wants a better world and he won't use a shitty flip phone to do it but I think that with some direction and restructuring this project can become something more.
<Ke>
I think shitty flip phone might be offensive, many people actually like that form factor
<Hodges>
I mean to say that he literally cannot use anything but a flagship Android device.
<Hodges>
And that moving from one platform to another will obviously curtail some growing pains
<Hodges>
But some are too great to overlook.
<Hodges>
We need to protect people from themselves without them ever knowing it
<Hodges>
But we can't step on any toes
<Hodges>
This is just one of many such initiatives for pushing innovation http://alphalabgear.org/
<Hodges>
Or well, hoarding it if acquired which is some peoples' hope.
<Hodges>
This project has a frightening lack of focus for such a serious topic. I'm ready to be scolded, reprimanded, or have my intelligence/savvy attacked while fielding any sort of demerit anyone here can think of and I'd take it all for truth.
<Hodges>
Someone - Tell me to shut up or egg me on. I'm 7 beers in, I'm heated, I'm red blooded, I want to fight, because I'm a goddamn freedom loving American.
<GTHodges>
I'm heading off to bed. If anyone hsd any musings I'll see them in the morning.
Hodges has quit [Client Quit]
<sn0wmonster>
luke-jr, didn't mean to imply that, just that the isolation model of Neo900 provides control of it that I don't believe Librem5 has.
chomwitt has quit [Ping timeout: 248 seconds]
<GTHodges>
sn0wmonster: What do you see as being the limiting factor in adopting another's existing device, or perhaps even several to make them into a neo900++
<Joerg-Neo900>
I'll ned to think a bit to give a _good_ concise answer to >>How does the Purism phone compare to neo900?<< - beyond the obvious formfactor dufference
Hodges has joined #neo900
<Joerg-Neo900>
there's a lot where those projects and devices differ. Probably one most remarkable one is the philosophy
<Hodges>
I agree.
<Hodges>
I truly believe that is the problem.
<Hodges>
And I don't see this project successful, however pure it's intentions, the talents of the men behind it, with any amount of hours ever succeeding without it being rooted in philosophy.
<Hodges>
1, Why - This is the core belief of the business. It's why the business exists.
<Hodges>
2. How - This is how the business fulfills that core belief.
<Joerg-Neo900>
purism propagates / implies a - in my book false - belief that you could mitigate rogue software by a hardware switch. In my book when you got rogue malware on your system you're doomed no mater what and your best option is to keep your system clean by being your own sysop with root access that you don't grant indiscriminately to any manufacturer or app installer
<Hodges>
3. What - This is what the company does to fulfill that core belief.
<Hodges>
I think their approach is all bullshit.
<Hodges>
I'm not as savvy or privvy in that field but I just feel it inmy gut.
<Hodges>
Continuing: 1. Step Back and Use "Why" to Think About Your Own Business
<Hodges>
2. Incorporate "Why" into Your Marketing Copy
<Hodges>
3. Redefine your Buyer Personas
<Joerg-Neo900>
Neo900 is clearly focused on Maemo5 fremantle which has proven to work fine on smartphone. Librem wants to create their own PureOS based on debian, but without any preexisting phone background
<Hodges>
I'm relieved to hear that.
<Hodges>
Because I think it reaching market will just set the market back.
<Hodges>
I feel the only way you can better people on such a philosophically rooted position is to just do it for the without their knowledge.
<Hodges>
Hell, even apple has done so with OTR communication
<Hodges>
I want the Neo to be true to itself, and true to it's customers but they need to believe they want it.
<Hodges>
Unless you want this product to not reach mass appeal
<Hodges>
Have a dozen revisions of increasing order and scale
<Hodges>
Or get a large corporate backing from a security firm that'll provide generous seed money with no unfair expectations outside customer support.
<Joerg-Neo900>
Hodges: you should take that nap you planned ;-)
<Joerg-Neo900>
I will be still around when you wake up
<Hodges>
Haha, my bad. I have a lot going on right now.
<Hodges>
I just quit my dream drop and went on vacation for 8 months
<Hodges>
dream job*
<Hodges>
And just never realized I had something I felt I stood the smallest chance of contributing too in a self-fulfilling near-spiritual way that's also a fuck-you to the gov't and every other money grubby corporation
<Hodges>
I hope that when we next talk it's a bit more focused.
<Joerg-Neo900>
:-)
<Hodges>
Let me know if you'd be interested in some kind of other secure form of communication.
<Hodges>
Because I am an idiot but I'm an endearing one to enough people that I might be able to get you at a minimum, stickers from google or free stuff from Razer
<Hodges>
Or at best, 1000s of Nokia N900s
<Joerg-Neo900>
let me get my coffee first, I'm not awake yet
<Hodges>
Tonight or some other time works for me. It's 6:45 am and I've just moved so my days have been spent puppy proofing the yard and repairing 30yrs of neglect in the yard.
<Hodges>
I'm tired but awake for the first time in months but I trust your time and judgement completely.
<Hodges>
LMK
<Joerg-Neo900>
1000 N900 sounds good and useful anyway
OppositeDay has joined #neo900
OppositeDay has quit [Client Quit]
<Hodges>
I've been fighting pulling the trigger on one now.
<Joerg-Neo900>
and I'd like to see you helping with kickstarter
<Hodges>
Let me know where you stand with potentially chatting with a nut.
<Hodges>
What timezone are you?
<Joerg-Neo900>
JR timezone virtual - aka JRTZV. It has 27h per day
<Hodges>
DDG doesn't give me any hits and I don't get the joke, lol
<Hodges>
Are you saying you are a cannuck?
<Hodges>
you don't have enough sunlight?
<Joerg-Neo900>
I'm just telling you that my local everybody's timezone isn't in sync with my very own one
<Hodges>
You upwork much? haha
<Joerg-Neo900>
my own timezone is ~27h per day but with huge jitter
<Joerg-Neo900>
my PC thinks it's 12:55 noon and I just have my first coffee
<Hodges>
I still don't follow. I feel like I've only had one day in my life.
<Joerg-Neo900>
anyway let me continue on >>How does the Purism phone compare to neo900?<< - One big difference is we believe in 100% open hardware, if the user can't get free public datasheets and schematics for every component in their device, they don't really own it
<Hodges>
Yes.
<Hodges>
Continue..
<Joerg-Neo900>
I'm also trying to involve community into development as much as possible: see our documentation that's available from very beginning and community can watch it evolve and can contribute
<Hodges>
FYI: 'm keeping all of this and my personal 'Sitcky Note' web-app.
gareth__ has quit [Ping timeout: 252 seconds]
<Hodges>
One immediate suggestion: BookStack and lots of well done D3JS like graphics
<Joerg-Neo900>
and of course based on our philosophy that the mere linux system has to be under exclusive control of the device owner and must not get tainted by any malware, we rely on an untainted linux to defend itself against threats and we provide all hw means you could think of (and some you never could think of before you seen them on Neo900) to facilitate that. Particularly we got a unique modem sandboxing and monitoring and we protect your bootloader
<Joerg-Neo900>
against evil maiden attacks
<Hodges>
Bookstack is a modern self-hosted wiki that is easily and freely contributed too
<Joerg-Neo900>
yes, we urgently need our own wiki
<Hodges>
For really making a proper 10-point plan to prime the technies like use to break their dogmatic ways
<Hodges>
Or maybe just see it as an easy thing to get into
Oksanaa has joined #neo900
<Hodges>
But also frame it in a way that inspires and incites the majority.
Oksana has quit [Ping timeout: 248 seconds]
<Hodges>
Because I can just spiel out inappropriate facts about surveillance that make me sound insane.
<Hodges>
with a few too many beers of course, and it's all true.
<Hodges>
The regular people need in on this too
chomwitt has joined #neo900
<Hodges>
Because we're not going to be the only ones hurting if things keep as they are
<Hodges>
The problems with neurology and the study of the genome is that none of these companies share information with each other
<Hodges>
Who knows how long healthcare has been pushed back
<Hodges>
Ridiculous software patterns are no different
<Hodges>
Most sickening are patents on basic concepts like nano-whatever is built is patented
<Hodges>
when there is no other way to do it in existence and it's just a building block of the world that can't be attributed to anything
<Hodges>
Sofware has been just as shitty and is turning around
<Hodges>
But the hardware however good it is, is just building on top of other shit it can only ever think it trusts
<Hodges>
Because of back doors, unknown bugs, unsafe designs electrically or physically, etc#
<Hodges>
And before getting into privacy or the lack thereof I think this project is on the right side of the fence now
<Hodges>
So i'm more than willing to dip out now if you'd like
<jonwil>
I am sad that I am (temporarily) using
<jonwil>
I am sad that I am (temporarily) using a phone that is totally insecure
<jonwil>
which is why I am using it for as little as possible (in this case phone calls and a clock so far)
Konsieur has quit [Remote host closed the connection]
Kabouik has joined #neo900
<jonwil>
Ok hmmm, turns out this thing is an alcatel and not an LG
<jonwil>
like I thought it was
<jonwil>
Another case of the carrier wanting to obscure the actual OEM that made the phone
<Joerg-Neo900>
almost as bad as Sagem ;-)
merlin1991 has quit [Remote host closed the connection]
<Joerg-Neo900>
Sagem and alcatel are certainly the most annoying two brands of phone U ever touched, UI-wise
<Joerg-Neo900>
I*
merlin1991 has joined #neo900
<jonwil>
Looks like its got a MediaTek MT6580M SoC, no clue if that's evil like the Qualcomm parts or just junk.
<Joerg-Neo900>
hmm, MT isn't all that bad iirc
<Joerg-Neo900>
though... SoC, well....
<Ke>
I think the MT devices without baseband might have been ok
<Ke>
MT8173C was investigated for chromebooks
<Ke>
or libreboot
<jonwil>
Its only got a lowly 4GB of internal storage (plus a MicroSD slot)
<jonwil>
Yeah this definitely has baseband on-board (Google seems to confirm it plus there is no way they could do a phone for $70 and have a separate baseband chip)
<Joerg-Neo900>
well, in this particular case rather "without APE" than "without baseband" ;-)
<Joerg-Neo900>
it's the typical "economic" design then, that runs APE in baseband processor
louisdk has joined #neo900
<jonwil>
well its quad core
<Joerg-Neo900>
then they might have assigned one core exclusively to BB
<Joerg-Neo900>
depends on config
<Joerg-Neo900>
no matter what, it's the "shared RAM, shared CPU" design
<jonwil>
Yeah
<jonwil>
As I said, I dont care so much since its just a cheap crap thing to keep me going until I decide what to do
<jonwil>
and I wont be doing anything on it that I care about
<Joerg-Neo900>
:-)
<jonwil>
I use it for phonecalls and for SMS and for clock only so far
<Joerg-Neo900>
technically there's nothing wrong with that design, until you start thinking about security
<jonwil>
At least Alcatel seem to publish something that looks vaguely like a kernel source tree (which is more than I can say for some OEMs)
<Joerg-Neo900>
Wow!
<jonwil>
Not that I actually intend to do anything with that code
<jonwil>
But at least I know my money didn't support a company that profits from violating the copyright of others.
ravelo has quit [Quit: Connection closed for inactivity]
chomwitt has quit [Ping timeout: 246 seconds]
jonsger has joined #neo900
chomwitt has joined #neo900
jonwil has quit [Quit: ChatZilla 0.9.93 [SeaMonkey 2.48/20170707010522]]
gareth__ has joined #neo900
chomwitt has quit [Ping timeout: 246 seconds]
louisdk has quit [Ping timeout: 255 seconds]
jkepler has quit [Remote host closed the connection]
freemangordon_ has quit [Ping timeout: 240 seconds]
ravelo has joined #neo900
<jackhill>
Joerg-Neo900: thanks! It it difficult for me to read the marketing material and know what is really going on.
<Joerg-Neo900>
Neo900 even offers two levels of evil maid protection, the 'normal' (default) one being effective against the more common and more dangerous "evil USB charger, at airport" scenario where you use an unknown public /rogue) USB charger that reflashes your bootloader or reads out your data. The more effective "full protected mode" is impossible to revert by any attacker that has no full featured hardware rework and forensics lab with them, and
<Joerg-Neo900>
sufficient time to accomplish the attack (more than an hour, possibly several hours of time)
jkepler has joined #neo900
<Joerg-Neo900>
plus for the latter, you can add mechanical safeguards like seals that would show you an attack attempt when broken
<Joerg-Neo900>
since those seals live under the battery cover, they are not possible to get investigated and forged beforehand by any attacker, increasing the number of access opportunities needed by attacker by one
<Joerg-Neo900>
nice sidenote comment: we achieved most of that by insignificant or completely no impact to BOM, just by smart design
<jackhill>
Joerg-Neo900++ thanks for your hard work
<Joerg-Neo900>
in short: Neo900 is designed from scratch with security in mind
<Joerg-Neo900>
hardware kill switches are considered an inapt means to increase security/privacy, we evaluated this quite thoroughly
<Joerg-Neo900>
they are a nice-to-have (though actually not _so_ nice when they are mechanical and not under control of the considered-safe OS) but they don't suffice to establish any reasonable level of security
jcarpenter2 has joined #neo900
<Joerg-Neo900>
the 3 topmost design goals for Neo900, in sequence of decreasing priority: full control of user over the device, Openness, pretty good security
<Joerg-Neo900>
hmm, that wording :-/ let me retry: user's full control over the device
jcarpenter2 has quit [Ping timeout: 240 seconds]
<Joerg-Neo900>
this obviously contrasts to pretty much all pther companies' approaches where "leete design" and/or "merchantability" and or "ROI" are the topmoost design goals
<Joerg-Neo900>
even shorter: "other companies: `trust US for your device security´ - Neo900: `trust your own evaluation of what you get from us´"
<Joerg-Neo900>
for that we provide al the needed documentation so you actually _can_ do that evaluation beforehand, users even _did_ contribute during design phase to improve the design
<Joerg-Neo900>
I rhink the original question of >>How does the Purism phone compare to neo900?<< is covered now? Any more questions regarding that?
<Joerg-Neo900>
I think Purism is doing a nice job designing a modern open (to whatever extent) linux smartphone, and I really hope they succeed. However their design goals and concept differ from ours, they follow a different approach
<jackhill>
Joerg-Neo900: I think that covers it, thanks.
<jackhill>
I was sort of surprised to see that they had so much interest compared to Neo900, so I was curious what the difference was.
<jackhill>
(I'm more used to the PC market where everything follows a very similar architecture (untrustworthy as it is))
<Joerg-Neo900>
Neo900 is 'old', bluntly said. We can't do "modern leete octocore" as Purism does. for a number of reasons, once being availability of SoCs and docs for Socs, another being our goal to provide 100% maemo compatibility to the existing app base and OS, without any need to even recompile a single app
<Joerg-Neo900>
for Neo900 that step to go "modern and leete" is planned for the Next device: Step2
<Joerg-Neo900>
so while Purism basically starts at square#1 with their PureOS, we want to continue where we were left behind by Nokia with N900/maemo-fremantle
<Joerg-Neo900>
I've seen 2 dozen similar projects given birth during last maybe 10 years, and more than one dozen failed due to exactly the (lack of working) software ecosystem
<Joerg-Neo900>
maemo is a mature working (though meanwhile slightly dusty) software ecosystem
<Joerg-Neo900>
it's based on Debian just like PureOS plans to be
<Ke>
Joerg-Neo900: I think the difference now is that we have a working kernel, working debian and working compilers
<Joerg-Neo900>
we have a working everything basically, including a "development platform" called N900
<Ke>
I am not sure I would like to have a non general purpose operating system on my phone
Pali has joined #neo900
<Joerg-Neo900>
yes, no real hacker wants that ;-)
<Joerg-Neo900>
maemo is as general purpose as it gets
<Ke>
why not just package maemo sw on debian?
<Joerg-Neo900>
err, because it's a different distro. Why not package mint on ubuntu?
<Ke>
because debian has wider developer base than maemo?
<Joerg-Neo900>
sorry, can't follow
<Ke>
well doesn't matter
<Joerg-Neo900>
a *lot* of maemo packages were just imported from debian proper, slightly adapted makefile if needed at all, and recompiled for maemo
<Joerg-Neo900>
obviously for GUI apps you want to adapt the GUI to accommodate Hildon desktop
<Joerg-Neo900>
but I've seen genuine openoffice running on maemo
<Joerg-Neo900>
of course you need a stylus or better even a mouse to operate that
<Joerg-Neo900>
there's a very huge and common misconception about how much maemo differs from plain vanilla devian. Hint: basically not at all, as far as anything related to apps is concerned, and not much regarding core system, and there are no blobs in core system at all. there are a few middleware things like liblocation and a few other dbus-attached daemons that were made proprietary by Nokia. however basically all of those got REed meanwhile and none of
<Joerg-Neo900>
them is necessary for maemo working like any other debian
<Joerg-Neo900>
debian/devuan*
<Joerg-Neo900>
there are optimizations like using upstart instead of sysv-init and not using an initrd, kust like you find them in a lot of derived distros, but basically maemo == debian
<Chris__>
What is about changing to devuan base?
<Joerg-Neo900>
changing to devuan basically kust means staying with debian proper, as it's debian that changed while maemo obviously didn't
<Joerg-Neo900>
maemo depends on cgroups which are shanghaied/taken hostage by systemd in "modern" debian
<Joerg-Neo900>
systemd forces a lot of changes into other packages and devuan is about removing those to keep compatible to "classical" debian
<Chris__>
We at Gentoo don't need that, we have OpenRC ^^
<Joerg-Neo900>
:-D
<Chris__>
And eudev. So a fork of system-owned-udev.
<Chris__>
*systemd
<Chris__>
So devuan is very welcome for maemo, yes?
<Joerg-Neo900>
indeed gentoo was one of the first distros that was brought up and working on N900 as alternative OS
<Joerg-Neo900>
yes
<Chris__>
Oh, really?
<Chris__>
Interesting.
<Chris__>
The anti-lennard-systemd-refugees.
<Ke>
I'm ok with systemd *ducks*
<Joerg-Neo900>
you can install debian incl systemd on N(eo)900
<Joerg-Neo900>
but I think it will perform sub-par
<Ke>
but it doesn't hurt to have competition, so I wish no harm to the alternative projects
<Chris__>
It's not about that systemd is not okay, it's just about making it to monopol. Unix is about choice, so Linux.
<Chris__>
Right.
<Joerg-Neo900>
fill ACK
<Joerg-Neo900>
full*
<Chris__>
We at Gentoo just support both.
<Joerg-Neo900>
for maemo there's no choice, since maemo needs cgroups
<Chris__>
And systemd don't supports it.
<Joerg-Neo900>
exactly, or rather it occupies them
<Joerg-Neo900>
or even more generically spoken, maemo has its own init system that conflicts with systemd
<Chris__>
OpenRC can.... use any init, runit for example.
<Joerg-Neo900>
so maemo needs a "debian" base that _allows_ alternative init systems
<Joerg-Neo900>
and that debian base is devuan
<Chris__>
Yes, binary distribution makes more sense for a phone.
<Joerg-Neo900>
note however that Neo900 the hardware is completely agnostic to any such OS considerations, you can run whatever OS you like on Neo900
<Joerg-Neo900>
we just provide all the documentation needed so you actually _can_ make your favorite OS work on that platform. while usually comapniies provide an OS and no docs how to port other alternative OS to their hw platform
<Ke>
that's definitely the way hacker phones should be
<Ke>
also unbrickable
<Joerg-Neo900>
that's what we thought :-)
<Chris__>
Yes.
<Joerg-Neo900>
yes, Neo900 like N900 is unbrickable
<Ke>
N900 does usb reflash mode or something?
<Ke>
from hw
<Joerg-Neo900>
unless you engafe the aforementioned "full protected mode" on bootloader
<Joerg-Neo900>
yes, N900 has ROMBOOT from USB in its boot-order-sequence. On Neo900 you can change that on a hw basis, "irreversibly"
<Joerg-Neo900>
<tongue in cheeck> Librem5 hw killswitch for disabling coldflashing? </tongue in cheek>
<Ke>
isn't ROM write protection quite a normal thing to have?
<Ke>
librem-5 probably won't have bootrom though
<Joerg-Neo900>
I don't know details about the Librem i-MXn SoC
<Joerg-Neo900>
there's no ROM write protection (used) in N900
<Ke>
in general that would be my preferred way of doing secure boot
<Ke>
like chromebooks do it
<Joerg-Neo900>
the point being that OMAP has an immutable ROMBOOTloader that can load from USB to RAM and execute
Kabouik has quit [Ping timeout: 240 seconds]
<Ke>
sure, RK3399 also has hw usb mode with open source tool support
<Joerg-Neo900>
N900 is a OMAP HighSecurity device that just doesn't engage / keep secure mode for secure boot. With N9 and meego / maemo6/HARMattan this changed. Neo900 can't and won't be a HS device since we mere mortals only can get hold of OMAP GP (general purpose) SoCs without the TrustYone/M-Shield stuff available. We basically just don't have access to the root certificate
ArturShaik has quit [Ping timeout: 240 seconds]
<Joerg-Neo900>
anyway secure boot is mainly usefil for manufacturers to deprive users/deviceownersfrom complete control over their own system. All other alleged benefits of secure boot can get accomlished by more "classical" means like OS provilege handling as well
<Joerg-Neo900>
there's no *real* dufference between employing TrustZone and implementing same functionality into your own bootloader and kernel on a "userland level"
<Joerg-Neo900>
both have vulnerabilities too, so no difference in that regard either
<Ke>
assuming your trust zone limits access to bootloader flashing, sure
<Joerg-Neo900>
TrsutZone is TI's (ARM's?) implementation of trusted computing on hardware level, dividing the systeminto a secure world and a untrusted world
<Joerg-Neo900>
I think TrusZone is ARM and M-shield is TI's implementation
<Joerg-Neo900>
and of course kernel can limit access to flashing kernel's imahe in flash, that's what kernel is made for
<Joerg-Neo900>
;-(
<Joerg-Neo900>
err :-)
<Ke>
but I don't want the same kernel limiting the flashing that does eg. tcp etc.
<Joerg-Neo900>
so as long as your root password is kept safe and your kernel has no vulnerabilities, a secure boot enabled system has no real security advantage over the regular linux
<Ke>
I don't quite agree
<Ke>
well ok, with your assumptions yes
<Ke>
but kernels do have vulns and I post my root pw on my homepage
<Joerg-Neo900>
then otoh your TrustZone / secure boot enabled device may have vilberabilities that can't even get fixed by you since you have no access to the secure world code
<Ke>
on chromebooks you can
<Joerg-Neo900>
not to all
<Ke>
since the secure boot is implemented with write protect
<Joerg-Neo900>
anyway, moot discussion. Unless you're a very big player you can't get access to HS SoCs
<Joerg-Neo900>
at least for OMAP
<Ke>
that's why having write protected rom would be awesome, but I can see why it could be hard to fit in
<Joerg-Neo900>
just saying we don't suffer severe disadvantages from that, actually I wouldn't want HS even if we could get it
<Ke>
SoC hardcoded secure boot is nasty yes
<Joerg-Neo900>
"write protected ROM" is a pretty fuzzy term. You probably *can* write-protect pages on OMAP NAND
<Ke>
though I would assume you could implement a SoC where the master key would be fused later on
<Ke>
yes
vlitzer has joined #neo900
<Ke>
does Neo900 ship with rom?
<Joerg-Neo900>
define rom!
<Ke>
512MiB NAND listed, I guess it can be used as boot source?
<Joerg-Neo900>
I'm EE, for me a ROM per definition is wrzet protected
<Ke>
yes
<Joerg-Neo900>
yes, it IS used as boot source
<Joerg-Neo900>
write*
<Joerg-Neo900>
on OMAP you can define boot order sequence. It may include UART, USB, NAND and MMC
Kabouik_ has joined #neo900
<Ke>
define as in fused in or jumper?
<Joerg-Neo900>
on N900 it has USB before NAND, for Neo900 you can revert that sequence so an atacker has no means to override an intact working bootloader in NAND
<Joerg-Neo900>
humper
<Joerg-Neo900>
jumper even
<Joerg-Neo900>
so NAND based bootloader could set write-prtect on its own (and kernel's) NAND storage pages
Hodges has quit [Ping timeout: 260 seconds]
<Joerg-Neo900>
and attacker has no means to override or bypass that NAND bootloader, e.g. via coldflashing aka ROMBL loading from USB
Kabouik- has joined #neo900
<Ke>
I guess your trusted bootloader could set the NAND protection on even without the jumper
<Joerg-Neo900>
IOW you could effectively brick your device then, when you manage to mess up the NAND bootloader so it doesn't biit the system but still will get executed
<Joerg-Neo900>
of course
<Ke>
obviously you need some recovery solution
<Ke>
eg. bootloader checks, whether you are pressing some button
<Ke>
if you are, then it allows flashing
Kabouik_ has quit [Ping timeout: 240 seconds]
<Joerg-Neo900>
we have a "hw switch" (remove battery cover) to allow boot from USB, and we have a drill-out trace to disable that "hw switch"
<Ke>
that's probably ideal solution
<Joerg-Neo900>
yiur botloader then can ask for a password before starting any flashing, and unless you can iverride that by removing battery cover, you're doomed if that mechanism fails
<Joerg-Neo900>
Neo900 doesn't ship with the override by batterycover disabled, but users can irreversably do it
<Joerg-Neo900>
I'm glad you think we found the ideal solution :-)
<Joerg-Neo900>
as said above: Neo900 even offers two levels of evil maid protection, the 'normal' (default) one being effective against the more common and more dangerous "evil USB charger, at airport" scenario where you use an unknown public /rogue) USB charger that reflashes your bootloader or reads out your data. The more effective "full protected mode" is impossible to revert by any attacker that has no full featured hardware rework and forensics lab with
<Joerg-Neo900>
them, and sufficient time to accomplish the attack (more than an hour, possibly several hours of time)
<Joerg-Neo900>
it's very hard to guarantee *absolute* security, as it would involve stuff like explosives inside chip cases etc, but I think we did what we could and reached a very good level of security sufficient for all but the most demanding requirements (like nuclear baseball)
<Joerg-Neo900>
usual encryption of all storage is up to user/admin, and we provide the hardware means to make boot sufficiently tamper-proof towithtstand attacks shorter than 1h
<Joerg-Neo900>
and then you still got plenty of cabaries in the system to tell when somebody removed the battery
<Joerg-Neo900>
canaries*
<Joerg-Neo900>
even with sytstem powered down
<Joerg-Neo900>
equivalent to the usual "cabinet opened" flag in BIOS of a regular PC
<Joerg-Neo900>
and again, all this didn't cause any increase in BOM cost
<Joerg-Neo900>
there are other security hradened phones with faintly same featureset (or less) regarding security, they sell for a 2k to 5k
jonsger has quit [Ping timeout: 240 seconds]
louisdk has joined #neo900
louisdk has quit [Ping timeout: 248 seconds]
ravelo has quit [Quit: Connection closed for inactivity]
ravelo has joined #neo900
pagurus has joined #neo900
chomwitt has joined #neo900
jonsger has joined #neo900
jcarpenter2 has joined #neo900
Hodges has joined #neo900
chomwitt has quit [Ping timeout: 255 seconds]
xmn has joined #neo900
chomwitt has joined #neo900
jonwil has joined #neo900
chomwitt has quit [Ping timeout: 255 seconds]
jkepler1 has joined #neo900
jkepler has quit [Ping timeout: 246 seconds]
jkepler1 is now known as jkepler
jabawok has quit [Ping timeout: 248 seconds]
Chris__ has quit [Ping timeout: 248 seconds]
Hodges has quit [Ping timeout: 258 seconds]
Hodges has joined #neo900
Hodges has quit [Read error: Connection reset by peer]
ravelo has quit [Quit: Connection closed for inactivity]