<mvaenskae>
bencoh: i thought they finally realized sharing main memory with the baseband was a bad idea
<mvaenskae>
but then just a simple code bug D:
<bencoh>
same here :)
<mvaenskae>
welp, let them have their wonderful android phones :)
<mvaenskae>
i wonder how many blackphone users will just install their usual goto apps and thereby render the blackphone idea useless anyways :)
<bencoh>
they'll probably end up with whatsapp, or telegram for the more adventurous :p
<DocScrutinizer05>
sorry for silence, still sick
<bencoh>
aww, get well :)
<mvaenskae>
DocScrutinizer05: i hope you will be getting better soon :o
<DocScrutinizer05>
yeah, I'm arguing with that damn virus every day. Also with my own evil mood that blocks me from e.g. contacting at least a dentist, so no more toothache on top of the flu
<mvaenskae>
what are you battling?
<mvaenskae>
bencoh: i was more thinking along the lines of the flashlight app that requires full system permissions ;)
<DocScrutinizer05>
flu with a nice little tooth problem in the company. Or the other way around
<mvaenskae>
D: not cool
<bencoh>
mvaenskae: :]
<DocScrutinizer05>
never know if my headache and paralyzed brain is caused by tooth or fever
<bencoh>
sneaky
<mvaenskae>
bencoh: aka google services framework ;)
SylvieLorxu has joined #neo900
<DocScrutinizer05>
(blackphone) I'm not sure the default install/config allows installing android apps from those common appshops
<DocScrutinizer05>
but the bug is pathetic
<mvaenskae>
that might be true but iirc you can get root-perms on it and thereby installing another app-store should be possible
<DocScrutinizer05>
sure
<DocScrutinizer05>
the silent circle guys obviously thought security concerns start outside of own system
<mvaenskae>
type confusion happens quite often i feel... if there were smart compilers to warn coders of possible type errors...
<DocScrutinizer05>
I mean, a silly not-even-buffer-overflow but simple negligence to sanitize/rangecheck inbound data type flags (aiui)
Pali has joined #neo900
<DocScrutinizer05>
I admit twinklephone (that part not designed or written by me) also is highly, even extremely, vulnerable to shellshock type env var exploits, and I didn't notice the problem when reviewing it back when. Hey it's a 8(?) years in the past
<mvaenskae>
twinklephone?
<DocScrutinizer05>
the industry lab standard state of the art SIP softphone
<DocScrutinizer05>
used by many SIP providers to test their infra
<DocScrutinizer05>
and by a lot of callcenters, it seems
<DocScrutinizer05>
I helped with development, particularly with the scripting hooks
<DocScrutinizer05>
and with I18N
<DocScrutinizer05>
German translation is by me
<DocScrutinizer05>
some maybe 30% of scripting concept are by me
<DocScrutinizer05>
hah, reviews date back to 2005, so 10 years
freemangordon_ has quit [Quit: Leaving.]
<DocScrutinizer05>
real fun when IRC chanops go mad and refuse to listen to polite concerns about their ways to "moderate" and administrate channels
<DocScrutinizer05>
:-S
<DocScrutinizer05>
honestly, I never before heard that some chanop claimed he *needed* to use /msg chanserv akick *.<isp>
<DocScrutinizer05>
and then even reacting with an explicit STFU when trying to discuss better ways to handle the problem than just enforcing ban without exempts on a whole ISP network
<DocScrutinizer05>
suggestion to innocent users who happen to be on same ISP: "get a cloak!" OMG!
b1101 has quit [Quit: b1101]
b1101 has joined #neo900
<mvaenskae>
wait, some freenode chanop kicked a whole ISP out?
<DocScrutinizer05>
banned, yes
<mvaenskae>
that's pretty bad
<DocScrutinizer05>
yeah, and he banned me for telling him
<mvaenskae>
i hope he angers some other chanop in the process
<DocScrutinizer05>
I doubt, he's chan founder
<mvaenskae>
hmpf, may i ask what channel?
<DocScrutinizer05>
"I asked the VUA's and they said just 'thanks', so this discussion end here"
<DocScrutinizer05>
guess what channel ;-)
<mvaenskae>
ubuntu?
<DocScrutinizer05>
#debianfork and #devuan
<mvaenskae>
devuan?! are they mad? they need everyone interested in their cause
<DocScrutinizer05>
yes, of course
<DocScrutinizer05>
poor coagen is massively floored on the task
<mvaenskae>
welp, that's how to get people to use systemd-debian
<mvaenskae>
so let's maybe not promote devuan on the neo900 :)
b1101 has joined #neo900
<DocScrutinizer05>
it's a pity the VUAs don't care or even know much about IRC it seems.
<mvaenskae>
let's just go full blown gentoo, the easiest to manage anyways as we will need make everything from sauce ;)
<bencoh>
and emerge on device, yay
<bencoh>
(thx but no thx :p)
<DocScrutinizer05>
actually here's a more comprehensive copy of #devuan channel http://paste.opensuse.org/29416266 . The real action however obviously happened on #debianfork
<DocScrutinizer05>
first troll, first epic fail of chan administration. A week before they stated they don't need any help ;-)
<DocScrutinizer05>
*shrug*
<DocScrutinizer05>
o/ back to bed
<Wizzup>
mvaenskae: gentoo yay ;)
<mvaenskae>
bencoh: we can alwaus create a binhost on gentoo or distcc with crossdev ;)
<mvaenskae>
also it isn't too bad, just 6h for a kernel \o/
<Wizzup>
mvaenskae: that long? kernel compiles should be <1 hour on a 1ghz
<Wizzup>
I guess if you have slow storage...
b1101 has quit [Quit: b1101]
<bencoh>
Wizzup: on arm ?
<Wizzup>
yes
<Wizzup>
Maybe if you're compiling the debian kernel, then not
<Wizzup>
They're too '=m' happy
<mvaenskae>
Wizzup: i was refering to a hardened kernel with a couple of modules :)
<Wizzup>
But I have a lot of arm devices, many 1Ghz or less, and kernel compiles don't take too long
<Wizzup>
mvaenskae: ah, I do that much faster on a single 1ghz arm device
<DocScrutinizer05>
thermal throttling? I hope not :-o
b1101 has joined #neo900
<mvaenskae>
oh, pardon :o the ti dm3730 is armv7 :p
<mvaenskae>
:o
<mvaenskae>
i am a big dum dum for mixing that up
<DocScrutinizer05>
meh!
<mvaenskae>
DocScrutinizer05: it is armv7, correct?
<DocScrutinizer05>
I'm admiring people who got the faintest clue about that maze that is ARM labels
<DocScrutinizer05>
yes, correct
<mvaenskae>
Wizzup: i was totally wrong, pardon my lack of knowledge :)
<Wizzup>
:)
<DocScrutinizer05>
anyway as long as you **got enough RAM**, compiling should neither take ages nor eat that much energy so the device would heat up unreasonably
<mvaenskae>
is armv8 backwards compatible to armv7?
<DocScrutinizer05>
there my own clue about that maze already ends
<Wizzup>
mvaenskae: btw, if you think HAM was slow, then just think about emerge
<Wizzup>
It sounds nice for playing around / testing
<Wizzup>
Not for 'end users', if that's what you may be aiming at
<Wizzup>
(I actually have quite a few arm devices that are self-updating with emerge, using binhost, updates still take quite some time)
<DocScrutinizer05>
(v8) but I *guess* that v8 should know how to run v7 code
<DocScrutinizer05>
would be weird to drop opcodes, they usually don't do that for newer models in a series of CPUs
<DocScrutinizer05>
I mean, you frequently need those functions provided by those opcodes anyway
<mvaenskae>
good :)
<mvaenskae>
then i will wait for armv8 to arrive before acquiring another arm board (thoroughly disappointed by my pi)
<mvaenskae>
kerio: thanks though for showing me alix; i think their apu counterparts make great routers :o
<DocScrutinizer05>
Enhancement to the AArch32 functionality
<DocScrutinizer05>
Relatively small scale additions reflecting demand
<DocScrutinizer05>
Maintaining full compatibility with ARMv7
<Wizzup>
nice
b1101 has quit [Quit: b1101]
<mvaenskae>
DocScrutinizer05: thank you very much :) then i will wait for armv8 boards :)
<DocScrutinizer05>
yw :-)
<kerio>
mvaenskae: yep
<kerio>
they run openbsd :3
<mvaenskae>
kerio: i would put gentoo on it ;)
<kerio>
fu
<kerio>
:v
<DocScrutinizer05>
mvaenskae: actually I looked up ARMv8 board a few days ago and pasted some URLs in here
<mvaenskae>
DocScrutinizer05: that will make compiling for the neo900 much easier i think; especially as they will likely be multicore and therefore having much more performance than the neo900 :)
<kerio>
DocScrutinizer05: pls armv8 neo900
<mvaenskae>
kerio: what's bad with putting gentoo on it? i know gentoo the best :)
<kerio>
with 32gb of ram
<DocScrutinizer05>
*sigh*
<Wizzup>
mvaenskae: no one will stop you from doing that
<kerio>
mvaenskae: gentoo had more than 2 remote holes in the default install in a heck of a long time!
<Wizzup>
I may attempt it as well
<Wizzup>
ah, oft repeated openbsd fallacies ;)
<DocScrutinizer05>
mvaenskae: yes, a compiler should provide genuine armv7 target arch on a armv8 machine
<kerio>
Wizzup: nah we're talking about alix boards for routers
<Wizzup>
ahhh
<Wizzup>
ok
<kerio>
yes, "fallacies"
<mvaenskae>
kerio: mind pointing me to reports? i would like to check if i have unknown holes :)
<kerio>
except that i never get a link to a CVE for a remote-exploitable third bug
<kerio>
mvaenskae: it's mostly tongue-in-cheek
<Wizzup>
(exactly)
<DocScrutinizer05>
may I quote "i would like to check if i have unknown holes" ? ;-D
<kerio>
ROFL
<kerio>
i wasn't even thinking about that
<mvaenskae>
DocScrutinizer05: are you searching for the links currently? (else i will look for them) :)
<mvaenskae>
DocScrutinizer05: sadly they don't seem to be selling single-devices :/
<mvaenskae>
kerio: you can always remove it and it won't complain ;)
<freemangordon>
DocScrutinizer05: I know you're ill, but a question I think is important just popped in my head: have you ever though what bootloader Neo900 will use? as AFAIC one can't boot unsigned secondary loader on HS device
<freemangordon>
*AFAIK
<mvaenskae>
welp, the second link is seriously expensive D:
<DocScrutinizer05>
well, Neo900 is no HS device
<mvaenskae>
HS?
<freemangordon>
GP?
<freemangordon>
oh, ok
<DocScrutinizer05>
I just discussed that with vakkov yesterday
<vakkov>
today *
<DocScrutinizer05>
see, my brain is porridge
<mvaenskae>
kerio: you posting those low-power boards is really great though; finally i have a hardware source for possible router hardware
<mvaenskae>
freemangordon: what is "HS"/"GP" refering to? :)
<kerio>
mvaenskae: alternatively, buy a dreamplug
<DocScrutinizer05>
High Security, General_Purpose(?)
<kerio>
aka shitty armv5 board with two gigE
<DocScrutinizer05>
TrustZone, M-Shield, that stuff
<DocScrutinizer05>
also Aegis, sort of
<mvaenskae>
kerio: armv5? no kthxbai x)
<kerio>
but it needs to route shit
<kerio>
and that's it
<kerio>
the one i have is happily routing warez and pr0n to a usb HD via bittorrent
<kerio>
surely running a network is less taxing
<DocScrutinizer05>
freemangordon: right?
<freemangordon>
yep
<DocScrutinizer05>
3430 vs 3530, and 3630 vs 3730
<freemangordon>
I forgtot Neo900 wil use only the memory from N9, not the whole core
<freemangordon>
*forgot
<DocScrutinizer05>
:-)
<DocScrutinizer05>
yeah, N9 using OMAP3630, Neo900: DM3730
<DocScrutinizer05>
iirc
<Pali>
what is difference between omap and dm?
<DocScrutinizer05>
err, dunno, prolly nottin
<DocScrutinizer05>
DM* is the version for the plebs
<DocScrutinizer05>
or sth like that
<Pali>
so only different name
<DocScrutinizer05>
aiui yes
<DocScrutinizer05>
freemangordon: vakkov and me concluded RAM init is done in x-loader and might get locked after x-loader finishes. What's your info on that? Could RAM init get done again in NOLO/uBoot (3rd stage)?
<DocScrutinizer05>
(on HS)
<Pali>
nolo has code for ram init
<DocScrutinizer05>
\o/
<Pali>
but no idea if it will work in kernel/nolo which is not signed
<Pali>
err. I wanted to say uboot
<Pali>
not nolo
<DocScrutinizer05>
"our" NOLO isn't signed, you can patch it
<Pali>
we do not know if that code for ram init does not have to be signed
<Pali>
like L3 firewall configuration
<DocScrutinizer05>
actually I wonder if not even GP chips need a signed x-loader, just that the signing key is "public"
<DocScrutinizer05>
actually I guess ~50% of users in here could help
b_ has quit [Ping timeout: 250 seconds]
norly has joined #neo900
<DocScrutinizer05>
Pali definitely knows how to build powerkernel (original maemo kernel with augments)
<DocScrutinizer05>
freemangordon knows about building most recent kernel, sth I consider step two after making a as-genuine-as-possible maemo build
<timclassic>
Is there a good starting point on the Wiki? I would ultimately like to build as much from source as is possible, pulling in closed bits only as necessary. I've searched a bit and found the "Fremantle Porting Task Force."
<timclassic>
Not sure if that's quite what I need (yet)
<DocScrutinizer05>
not yet :-)
<timclassic>
Though I have a claim on a Neo900.
<DocScrutinizer05>
first step: get a "homegrown" maemo to work on N900
<DocScrutinizer05>
you'll need all the stuff you need to establish for that, when you later on build maemo for another target
<timclassic>
Sure, makes sense
<DocScrutinizer05>
I'm almost sure nobody outside of Nokia servers behind walls ever built a genuine maemo from scratch
<timclassic>
Wow, really? Everyone has just taken their packages and modified them?
<timclassic>
Hmmm
<DocScrutinizer05>
except probably stskeeps alias Carsten Munkert(?), see fptf reference
<freemangordon>
Pali: though, if you read to the end of that thread, there are some comments that this key is not that much useful
<Pali>
yes, I read it
<merlin1991>
iirc the factoid for sb images is
<merlin1991>
~#maemo SB
<infobot>
rumour has it, scratchbox is a cross-compiling system that uses binfmt_misc, rpc calls, and an nfs mount to make a cross-build appear to be 100% native, and is found at http://www.scratchbox.org/, hosted by maemo now. Also at http://maemo.merlin1991.at/files/SB
<freemangordon>
Pali: and yes, this is exactly the same key I have as ROOT.pem in my CSST installation :)
<Pali>
freemangordon: is there any tool or anything else which can extract keys & certificates from x-loader structure header?
<freemangordon>
I looked just briefly on that CSST thingie