<Turl>
<15mA peak current, 0.01-0.5 the power consumption of normal bt
<Turl>
according to wiki
<oliv3r>
read that ti blogpost; it's quite old though
<oliv3r>
it's a very good read
<oliv3r>
i'm half way :p
<oliv3r>
i think BLE should be conciderd for 'body worn sensors' etc
<oliv3r>
whereas dash7 is for anythign with bigger range
<oliv3r>
home automation probably doesn't fit well with BLE
<Turl>
yeah
<Turl>
it's actually being used that way (BLE)
<oliv3r>
though they do claim 100m for BLE
<Turl>
hadn't heard of dash7 until you mentioned it though
<oliv3r>
it's only 3 or 4 years old
<oliv3r>
there's af ully opensource stack; opentag
<oliv3r>
currently it's used mostly by the department of defense int he USA, airports etc
<oliv3r>
it IS being used commercially
FR^2 has quit [Quit: und weg...]
printallthething has quit [Ping timeout: 272 seconds]
printallthething has joined #linux-sunxi
earny has quit [Read error: Connection reset by peer]
earny has joined #linux-sunxi
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
newleaves has joined #linux-sunxi
earny_ has joined #linux-sunxi
earny has quit [Read error: Connection reset by peer]
newleaves has quit [Remote host closed the connection]
printallthething has quit [Read error: No route to host]
egbert has quit [Ping timeout: 246 seconds]
egbert has joined #linux-sunxi
HdkR has joined #linux-sunxi
earny_ has quit [Read error: Connection reset by peer]
earny has joined #linux-sunxi
earny has quit [Read error: Connection reset by peer]
earny has joined #linux-sunxi
earny has quit [Ping timeout: 252 seconds]
hipboi has joined #linux-sunxi
atsampson has quit [Ping timeout: 245 seconds]
atsampson has joined #linux-sunxi
[7] has quit [Disconnected by services]
TheSeven has joined #linux-sunxi
Gerwin_J has quit [Quit: Gerwin_J]
HdkR has quit [Remote host closed the connection]
YiWang has joined #linux-sunxi
pwhalen has quit [Remote host closed the connection]
pacopad has joined #linux-sunxi
kivutar has quit [Ping timeout: 245 seconds]
kivutar has joined #linux-sunxi
rz2k has joined #linux-sunxi
Quarx has joined #linux-sunxi
<libv>
aha, seems like puneet sent the schematic of his humming bird board
<libv>
ssvb: i could always go and disable overlays
<libv>
ssvb: but i really like the way you sneak in the overlay, and it's really good for performance
<libv>
aha, it's A20_PAD_STD
<libv>
by allwinner
<libv>
dated august 2013, while the a20_pad_std document we have is from april
earny has joined #linux-sunxi
nikrou has joined #linux-sunxi
<hipboi>
libv, we designed a board referred the A20_PAD_STD
<hipboi>
but the board was not bringing up
<hipboi>
after checking, we found there was error in the reference design
TomiK has joined #linux-sunxi
nikrou has quit [Quit: Leaving.]
kivutar has quit [Ping timeout: 248 seconds]
kivutar has joined #linux-sunxi
rz2k has quit []
YiWang has quit [Quit: Page closed]
<libv>
hipboi: so the later revision fixes this?
Fusing has joined #linux-sunxi
earny has quit [Read error: Connection reset by peer]
earny_ has joined #linux-sunxi
techn_ has joined #linux-sunxi
Faisal has joined #linux-sunxi
<Faisal>
hi there
<Faisal>
thank you all for your work
<Faisal>
im wondering, is linux-sunxi affiliated with allwinner ?
<libv>
Faisal: what do you mean with affiliated?
<libv>
Faisal: and why do you ask?
<Faisal>
libv: out of curiosity, just wondering if linux-sunxi is independant or if linux-sunxi releases software for allwinner officially (wondering if the linux-sunxi folks have access to all the hw documentation allowing to develop drivers)
<Faisal>
I need to run a command line everytime i reboot to get the bluetooth module up and running, the weird thing is that it doesnt work the first time i run it, but always works the second time : sudo /usr/bin/brcm_patchram_plus -d --patchram /lib/firmware/ap6210/bcm20710a1.hcd --enable_hci --bd_addr 11:22:33:44:55:66 --no2bytes --tosleep 1000 /dev/ttyS1
<Faisal>
. any ideas ?
<steev>
yes the WHOLE thing
<oliv3r>
in this case, this would be hte 'extraargs'
<ssvb>
the x86 guys developing 2d drivers are just spoiled by having insanely overpowered hardware (so that they can ship any junk driver and get away with that)
* ssvb
had to live with catalyst for about a month when experimenting with Open CL programming, and catalyst really sucks for 2D
<mnemoc>
ow
* mnemoc
ordered an amd apu a4-5000 laptop some days ago :\
paulk-aldrin has joined #linux-sunxi
earny_ has joined #linux-sunxi
earny has quit [Read error: Connection reset by peer]
<vrga>
ATI sucks on linux? Since when was this news? mid 2005? :p
Fusing has joined #linux-sunxi
<mnemoc>
hope tells you "things will get better!"
<vrga>
realism says "get nvidia, forget about it"
<vrga>
:p
<vrga>
(i know i know, shutting up)
rz2k has joined #linux-sunxi
<mnemoc>
:)
<vrga>
in other news, waiting for my funtoo rootfs compile the basic stuff is really annoying.
earny has joined #linux-sunxi
earny_ has quit [Read error: Connection reset by peer]
<ssvb>
vrga: nvidia is not perfect either, they sucked at implementing server side gradients, don't know if it got fixed now
<vrga>
on the other hand, performance might improve over the ubuntu core and its hilarity.
<vrga>
downside is, updating is done strictly offline...
<ssvb>
vrga: the linux distributions even patched the cairo library in order to disable the use of accelerated gradients because of the nvidia blob
<vrga>
ssvb, what are those?
<ssvb>
vrga: the things that are used, for example, in svg
<oliv3r>
ssvb: with catalyst you talk about how shitty the closed blob driver is right?\
<vrga>
( i cant say much for ati other than hearsay, having not used it, but i've never had much if any problems with nvidia either on windozers or linux)
<oliv3r>
so mnemoc no worries; you'll be using the oss driver mainliny anyway, and that's pretty descent
<mnemoc>
oliv3r: :)
<mnemoc>
oliv3r: I just hope at least my ssd arrives before fosdem :\
<mnemoc>
running from a usb stick stinks
<ssvb>
vrga: that's because the distributions and the applications take a "pragmatic" approach and avoid using any accelerated features that were ever bad on any hardware or driver version, leading us to the client side software rendering :)
<ssvb>
oliv3r: I'm talking how the 2d acceleration is on the decline in general, and you are not expected to get decent performance in x11 nowadays
<ssvb>
oliv3r: with all the blame being put on the "ancient 30 years old architecture which must die" instead of the actually shitty implementations of the drivers
<vrga>
its kinda both.
<ssvb>
how so?
Fusing has quit [Ping timeout: 260 seconds]
<vrga>
X is a dinosaur which you can barely even use in the network mode with modern DE's.
<vrga>
if you can even get it to work proper.
<ssvb>
this is just bullshit
<vrga>
what?
<vrga>
why?
<vrga>
VNC is not responsive enough for that anyhow.
<ssvb>
VNC has nothing to do with X11 network transparency
Fusing has joined #linux-sunxi
<vrga>
no, but when you get it reccomended as an alternative you get annoyed.
<ssvb>
let's say that there are proper applications, which work nicely with x11 network transparency
<ssvb>
that's mostly the old GTK2 based stuff
<vrga>
which are few and far between.
<ssvb>
and there are 'modern' applications such as Qt based, which often resort to the client side rendering, mostly because of the glitches in the poor x11 drivers
<vrga>
and there are no alternatives.
<vrga>
(to X that is)
* vrga
goes to check on the status of wayland
<ssvb>
to give one more example, Chromium sucks with x11 network transparency, but Firefox works fine
<ssvb>
there is still a choice
Quarx has quit []
<vrga>
ofc.
<vrga>
but there isnt really a choice in the backend tho.
<ssvb>
xorg failed because they don't have any sort of drivers "certification" process to clearly label shitty drivers as such
<vrga>
agreed.
<vrga>
gotta love the load average right now. Load avg: 19.5, 16.2, 10.1
<martin__>
Hi, is there anyone, who tried to make its own linux distro...following tutorial from http://linux-sunxi.org/Manual_build_howto stucks after starting a few subsystems. Another weird think, that linux print: Cannot find device "eth0" Bind socket to interface: No such device Failed to bring up eth0. While starting. Has anyone any advice ?
<mnemoc>
why the f* do they watermark such things as confidential???!
<itdaniher>
because they have more to gain from marking them vs not marking them
<mnemoc>
oliv3r: sure your boot.scr isn't forcing a fatload?
<mnemoc>
itdaniher: such thing is of no use unless people buys their A20
<oliv3r>
Turl: ext4load mmc 0 0x1 uImage works fine
<oliv3r>
but the boot.scr, which has exactly that line in it
<oliv3r>
fails
deasy has joined #linux-sunxi
<oliv3r>
/dev/sdd1 2048 67583 32768 83 Linux
<Turl>
oliv3r: don't you need sth like 0:1 to indicate first partition or sth?
<oliv3r>
ext4load mmc 0 0x48000000 ${kernel}
<Turl>
itdaniher: mirrored, thanks
<oliv3r>
Turl: if so, howcome if I manually do it, it works?
<oliv3r>
copy/pasting that exact line
<oliv3r>
** Unrecognized filesystem type **
<oliv3r>
i get that 3 times
<oliv3r>
i don't understand
<Turl>
oliv3r: sure they come from that line?
<bfree_>
probably from when it is loading the boot.scr and is trying fat also. "env print" is your friend (and you can run the parts up to where boot.scr is loaded then do it again)
<Turl>
that ^ :)
<bfree_>
by "do it again" I mean "env print" again
<bfree_>
(and you probably also get them from attempts to load uEnv.txt)
<mnemoc>
the default u-boot env DOES try to vfat before ext4, /boot vs /boot-less
<mnemoc>
and*
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
<oliv3r>
i used this crap as example; and while i thought things where off; it got merged; so should have been right
tomboy64 has joined #linux-sunxi
<oliv3r>
mnemoc: how did you configure 12gb, i can't get it above 8
<Turl>
mnemoc: he says he got 8G there and the apu has some magical 4G builtin
<Turl>
err oliv3r ^
<Turl>
but I think it's bs and he actually bought a 8G laptop :p
rZr has quit [Read error: Connection reset by peer]
<mnemoc>
8GB laptop would be nice too, the doubt of what I have today
<martin__>
Trying to debug, why system is not booting - Cubieboard2. Last what I see on serial console: [....] Starting system message bus: dbus. ok. What should be run after DBUS ? Should user login ?
<mnemoc>
can't find the post where i read it comes with 4GB "built-in"
<mnemoc>
s/doubt/double/
<Turl>
martin__: sounds like it's booting ok
<Turl>
martin__: but you don't get a login tty?
<Turl>
martin__: are you passing console=ttyS0,115200 on kernel cmdline?
<martin__>
Turl: yes, didnt get it
rZr has joined #linux-sunxi
<markvdb>
is there any sunxi or similar board with more than 2Gb ram that will easily run Debian?
<Turl>
cubietruck has 2GB iirc
<markvdb>
(and preferably powerful cpu)
<Turl>
I don't think there's anything with >2GB
<markvdb>
Turl: yeah
<markvdb>
I want to do limited image postprocessing for the diybookscanner.org project on an arm board
<martin__>
Turl: that should be in u-boot ?
<Turl>
martin__: yeah
<mnemoc>
i think only marvell supports >2GB currently
<Turl>
markvdb: do you need that much ram to scan images?
<markvdb>
not for the scanning, but for postprocessing (autocrop, ...)
<markvdb>
tesseract ocr
<markvdb>
image compression
<markvdb>
I'm trying to find a good tradeoff between performance and cost
<markvdb>
if postprocessing on the board is too expensive in terms of performance, then it will be limited to just triggering
<markvdb>
and I imagine an olinuxino A10 LIME would be a good choice
<mnemoc>
the lime has 512MB
<markvdb>
yup
<markvdb>
that's plenty for just triggering and a web interface
<markvdb>
but when it comes to more cpu and mem intensive tasks...
<markvdb>
Turl: then one would probably want a cubietruck at least, or maybe
<oliv3r>
how in the world is this possible
<Turl>
markvdb: well, you could make a vm on your pc with 2GB and see how all those ocr things interact memory wise
<markvdb>
a radxa rock or odroid-xu I guess
<markvdb>
Turl: that's a nice suggestion indeed
<oliv3r>
i boot my kernel from mmc, boot sda1, no problem, i can mount mmcblk0p2 no problem. i tell it that the root is ON mmcblk0p2; VFS: Cannot open root device "mmcblk0p2" or unknown-block(0,0): error -6
<oliv3r>
I really should just give up
<mnemoc>
mmcinfo ?
<mnemoc>
or it's the kernel whining?
<mnemoc>
rootwait ?
<Turl>
oliv3r: 6 is enixio (no such device)
<Turl>
try rootwait as mnemoc suggests
<oliv3r>
mmc being really slow? fair nuff; defualt is with rootwait
<Turl>
yeah it is
<Turl>
many people have issues if not rootwait'ing
<mnemoc>
sometimes initialization can be "late" just a couple of miliseconds and it's enough to make root mounting fail
<mnemoc>
so just rootwait
fredy has quit [Excess Flood]
fredy has joined #linux-sunxi
<oliv3r>
rootwaiting
<markvdb>
hm, I guess if I can live with something headless, the Mele M9 might be a good choice?
<oliv3r>
well i removed rootwait because i wanted to see why things where failing
<markvdb>
quad a7?
<martin__>
Turl: I don't know, why this happen. I make boot.cmd on first partition, then transfer it into boot.scr. In boot.cmd there ttyS0 is defined, but in uboot I can see only
<oliv3r>
markvdb: you won't get much sympathy for using a31 here, powervr makes most people here puke :p
<oliv3r>
markvdb: you'll end up with a device mostly relying on AW for kernel updates and support
<oliv3r>
atleast for now
<markvdb>
I see, thought there was a lot of progress reverse engineering their gpu
<Turl>
mali? yeah
<Turl>
but not powervr afaik
<markvdb>
I know about Mali
<markvdb>
Luc V told me personally :-)
<Turl>
he's in here, libv :)
<oliv3r>
libv: finally, finally i remade from scratch; but no tty
<oliv3r>
last thing init gave me was upstart-socket-bridge
<oliv3r>
Turl: http://paste.debian.net/77119/ why am I not getting a login :( this is the alip rootfs; it should 'just work'
<oliv3r>
i did init --verbose :p
<oliv3r>
since when has it become so hard to boot linux? for years and years; booting was as simple as having a proper /sbin/init
<oliv3r>
with this systemd and upstart crap .. really ... it's horrible
<Turl>
oliv3r: can you type stuff?
martin__ has quit [Ping timeout: 272 seconds]
<Turl>
or did it hang?
martin__ has joined #linux-sunxi
<oliv3r>
i can type stuff; it only echo's it
<Turl>
oliv3r: no inittab I suppose? :p
<Turl>
oliv3r: can you hook hdmi or something and see if you get a tty there?
<oliv3r>
Turl: ubuntu doesn't use inittab
<oliv3r>
it uses upstart!
<oliv3r>
and upstart should always start a serial console
<oliv3r>
i have sdd; with sda1 = fedora; sda2 = alip; sda3 = debootsrapped ubuntu saucy
<Seppoz>
you should even be able to debootstrap it arnt you?
<doneill>
actually it's single-cpu only so it won't work with the a20+s
<Seppoz>
oliv3r where are you stuck on it?
<oliv3r>
fedora runs great; alip i can boot, and if i put the console to the tty0 i get an error about the roofs being 'broken'; i can either wait, s to continue or m to 'fix' things; if i do s, i get a ro / and it 'works'. but ro. if I put the console on ttyS0 i don't get that and it just hangs
<oliv3r>
ubuntu gives the same, but with 'skip' it sort of half crashes, about /tmp not being available
<oliv3r>
it LOOKS like it doesn't like the empty fstab
<oliv3r>
but libv says his identical alip tar ball 'just works'
<oliv3r>
maybe the kernel doesn't like the ext4 somehow? incompatible ext4's? fsck doesn't notice anything strange