**** BEGIN LOGGING AT Wed May 28 02:59:59 2014
May 28 03:41:24 <ketas> i wonder how much one can exceed 1800mv or adc before adc input blows up
May 28 03:41:33 <ketas> must read specs
May 28 03:45:13 <foreverska> That doesn't seem like something that should be regularly tested.
May 28 04:07:00 <ketas> foreverska: was that to me?
May 28 04:08:00 <foreverska> You are the first one to talk in hours...
May 28 04:08:09 <raspberrypifan> nah uh
May 28 04:08:20 <foreverska> <.<
May 28 04:08:55 <foreverska> Uhhh ketas was the first message I saw since I sent that.
May 28 04:10:52 <raspberrypifan>  w/e
May 28 04:14:23 <vagrantc> i'm sure someone could write up a bot on a beaglebone black to populate the channel with chatter...
May 28 04:15:19 <foreverska> I wrote an IRC bot years ago.  It's just overgrown telnet.  A beagle could host a small crowd.
May 28 04:16:30 <ketas> i should dig up my old irc bot code
May 28 04:16:37 <foreverska> How you generate random chatter that begins to look real is beyond me though.
May 28 04:16:38 <ketas> hell, that did everything
May 28 04:19:27 <vagrantc> grab irc logs from a few weeks and run it through polygen...
May 28 04:22:51 <foreverska> Well sure if you wanted to do it the easy way...
May 28 04:48:54 <johnwalkr> has anyone successfully gotten glibc (instead of eglibc) working in recent debian?
May 28 04:49:17 <johnwalkr> they aresupposed to be binary compatible but i can not get a certain binary driver working under eglibc :(
May 28 04:49:25 <tbr> ask debain?
May 28 04:50:25 <johnwalkr> yeah i guess it’s not really BBB specfic, good idea
May 28 04:50:29 <thurgood_> I think it's valid here, since it is the official distro of the beagleboneblack  now
May 28 04:50:57 <johnwalkr> i read the history of why there was eglibc in the first place and it’s rally annoying
May 28 04:51:10 <johnwalkr> just because one developer didn’t like supporting embedded systems :(
May 28 04:55:46 <tbr> thurgood_: sure, though his chances to get an answer to a debian question, from debian people are much higher, aren't they?
May 28 04:58:35 <thurgood_> sure
May 28 04:59:35 <vagrantc> johnwalkr: why do you want glibc?
May 28 05:10:17 <thurgood_> does anyone know if there's a trick to getting sound working on android 4.3 with the bbb?
May 28 05:54:06 <Ben___> Hi, is anybody out there who has a (C/C++) toolchain for the new BBB debian?
May 28 05:55:35 <AE7IK> You mean can compile for one?
May 28 05:56:24 <AE7IK> I can log into my BBB...
May 28 05:56:54 <AE7IK> (And it's running the latest debian BBB image, though I have little room left on the device at this point in time)
May 28 05:57:14 <vagrantc> Ben___: apt-get install build-essential ?
May 28 06:08:31 <Ben___> vagrantc: already did this from debian multiarch howto, but where can I find now my Arm-gcc?
May 28 06:10:06 <vagrantc> Ben___: you want a cross-architecture toolchain?
May 28 06:11:05 <Ben___> vagrantc: YES, i have an 32 bit wheezy host (vm)
May 28 06:12:13 <Ben___> for angstrom distri there was an toolchain.tar.gz available
May 28 06:15:33 <vagrantc> multi-arch cross-toolchains are a work-in-progress...
May 28 06:15:51 <vagrantc> although the toolchains from emdebian might work fine for wheezy...
May 28 06:17:33 <Ben___> emdebian didn't work for me I get stuck somewhere (can't remember where exactly right now), but I got something promising right now from robertc nelsons github
May 28 06:18:42 <flufmnstr> so, i started playing with my BBB again. in cloud9 i ran the fade.js. when i stopped it the LED stayed in its last state(no biggie, it happens). but now i cant get anything to work on that pin. is there a pin reset command i can toss at it?
May 28 06:24:05 <johnwalkr> i lost my connection and foret who asked… but the reason I want glibc is that I have a propriestary camera driver. it’s binary only and the vendor won’t recompile it for me
May 28 06:24:25 <johnwalkr> and it segfaults no matter what i try using recent debian or ubuntu
May 28 06:25:05 <johnwalkr> I got it working with some ancient ubuntu but then the BBB support for everything else for that version is spotty
May 28 06:26:00 * vagrantc wonders how much a different camera would cost...
May 28 06:32:16 <johnwalkr> vagrantc i would pay $5000 for a camera that worked out of the box and met my requirements
May 28 06:32:26 <johnwalkr> trust me I looked into that option :(
May 28 06:49:14 <flufmnstr> uh....nevermind. it seems to have fixed itself....
May 28 06:49:21 <flufmnstr> dang gremlins....
May 28 07:13:22 <das> has someone dealt with Tuxera's FS implementations ?
May 28 07:14:20 <das> they advertised 3 times the speed on VFAT and 10 times on NTFS
May 28 07:16:05 <_av500_> for NTFS?
May 28 07:16:12 <woglinde> das try out and tell us
May 28 07:16:54 <das> av500: yup
May 28 07:16:59 <_av500_> you wont get 10x the speed on writing to SD or EMMC
May 28 07:17:10 <_av500_> and you are limited with speed on USB too
May 28 07:17:20 <_av500_> so, all jut theoretical
May 28 07:17:26 <_av500_> on the BBB at leastr
May 28 07:17:45 <jpfau> I mean, if the speed advertisement comes from reduction in I/O overhead
May 28 07:17:54 <jpfau> then there's at least some potential improvement
May 28 07:18:10 <jpfau> but that seems rather unlikely, especially for sustained I/O
May 28 07:18:18 <_av500_> indeed
May 28 07:18:35 <das> well of course as soon as any support such as eMMc/USB is involved the performance is invisible
May 28 07:18:37 <jpfau> and it's sustained I/O where it hurts the most anyway
May 28 07:19:06 <das> they sent me some stuff that uses an SSD to bench their implementation
May 28 07:19:30 <_av500_> on a PC?
May 28 07:19:46 <das> x86 embedded, basically a PC yeah
May 28 07:19:50 <_av500_> making VFAT fast is easy
May 28 07:20:02 <_av500_> jut cache the whole FAT in memory
May 28 07:20:07 <das> :P
May 28 07:20:27 <_av500_> which basically a PC does anyway these days
May 28 07:20:39 <_av500_> and the linux blockcache does too
May 28 07:21:28 <das> I'll try to find some time this week to confirm their results
May 28 07:21:44 <_av500_> +1
May 28 07:22:29 <_av500_> blog about it
May 28 07:23:16 <das> I'm younger than blogs, I'll reddit about it though
May 28 07:27:15 <_av500_> lol
May 28 07:30:23 <foreverska> Anyone know the max switching speed of a GPIO?
May 28 07:36:27 <_av500_> how fast do you need?
May 28 07:36:59 <_av500_> if you need fast, consider using the PRUs
May 28 07:37:16 <_av500_> there you can switch a GPIO at every 200mhz clock cycle
May 28 07:37:20 <_av500_> reliably
May 28 07:38:16 <foreverska> Oh I am.  I'm having trouble reading a circuit from there actually and when I thought about it it should easily read this.  The Hz function on my multimeter says the output is maxing at 6kHz.
May 28 07:40:05 <Rotti> hello
May 28 07:44:50 <foreverska> IRC can be a bit slow Rotti, if you have a question it's best just to ask it.
May 28 07:46:04 <Rotti> foreverska: i know ;) atm i don't have any questions... just came here to idle and read some interesting stuff
May 28 07:46:18 <foreverska> fair enough
May 28 07:46:33 <_av500_> foreverska: so you are using the PRU?
May 28 07:47:05 <foreverska> Yea.  I'm getting pretty good with the little buggers
May 28 07:47:58 <_av500_> good
May 28 07:48:04 <_av500_> blog about it
May 28 07:48:31 <foreverska> Currently working on my blog... Google Summer of Code
May 28 07:48:46 <_av500_> ah
May 28 07:50:37 <foreverska> You know... I kinda don't know why I asked earlier.  I actually wrote one pru to feed the other pru an ideal input.  If that works an external circuit should work if it's right.
May 28 07:52:14 <foreverska> I'm just not an EE guy.  I'm a pretty decent programmer but I'm largely lost when it comes to electronics.
May 28 07:52:38 <_av500_> just dont be afraid
May 28 07:53:10 <foreverska> lol I already fried a BBB since I started this project.
May 28 07:53:16 <_av500_> good :)
May 28 07:53:27 <_av500_> they are cheap
May 28 07:59:07 <foreverska> I do miss multiplication though.  You take floating point math for granted when you program C in userspace.
May 28 07:59:13 <johnwalkr> i fried 3!
May 28 07:59:15 <johnwalkr> :D
May 28 08:00:18 <johnwalkr> polarized power plugs are useless if you cna still plug them most of the way inthe wrong orientation ;)
May 28 08:00:48 <hype_> lol
May 28 08:02:26 <johnwalkr> wasn’t even on the bbb, but the power for a motor driver, which is eventaully connected to BBB via i2c
May 28 08:02:45 <KotH> foreverska: uhmm.. the PRU can do up to 200MHz, a multimeter will have trouble measuring that. but it will still show some bogus number
May 28 08:02:55 <KotH> foreverska: use a osciloscope instead
May 28 08:03:09 <_av500_> first rule when making a multimeter: always show something
May 28 08:03:26 <KotH> yeah
May 28 08:04:22 <foreverska> KotH: I wish I had oscilloscope money
May 28 08:04:52 <woglinde> foreverska you get a good enough oscilloscope for 200 euros
May 28 08:05:29 <foreverska> The multimeter's Hz function varies between 60 and 6000 directly corresponding with the knobs which control it.  It seems very right.
May 28 08:16:36 <KotH> foreverska: 2kchf for a tds2024 new
May 28 08:17:04 <KotH> foreverska: an anlog one is available for <100eur
May 28 08:18:04 <KotH> foreverska: and the multimeters frequency measurement is based on some assumptions about the waveform. if it deviates, you will have some value that varies with the input but is far away from the real number
May 28 08:19:32 <foreverska> Hmmm I really need my old hackerspace.  They had a Oscilloscope.
May 28 08:21:10 <foreverska> I guess adafruit has a $100 digital pocket Oscillo that can measure up to 100KHz
May 28 08:21:43 <Rotti> foreverska: what frequencies do you need? are analog signals required?
May 28 08:21:56 <Rotti> there are REALLY cheap logic analyzers from china
May 28 08:22:18 <_av500_> or use the PRU as one
May 28 08:22:47 <foreverska> The trouble with the pru is it's not registering Ons or Offs of the signal
May 28 08:22:58 <foreverska> and no I guess It's just digital
May 28 08:23:24 <Rotti> foreverska: then have a look at sigrok
May 28 08:23:42 <_av500_> check pinmux
May 28 08:23:48 <_av500_> check signal levels
May 28 08:25:29 <foreverska> Pinmux is right,  Feeding 3.3v to it turns the bit on.  The circuit fed 12v drives the PRU side to 330mv and the input to zero.  Disconnecting it from the 12v makes the output 3.36v and the logic at the PRU 1.  It just failes to switch when a signal is applied to the circuit.
May 28 08:29:27 <Rotti> foreverska: where are 12v? do you have any schematic?
May 28 08:30:23 <foreverska> not my schematic but it's built on these specs:  http://www.bgsoflex.com/v22/megasquirt_ShemV2.2.pdf
May 28 08:30:31 <foreverska> second page look for the input "COIL"
May 28 08:32:02 <Rotti> VCC is 3.3v?
May 28 08:32:11 <foreverska> Yes
May 28 08:32:37 <foreverska> atleast in my version it is.
May 28 08:32:52 <Rotti> and coil is 12v?
May 28 08:33:07 <foreverska> Ideally
May 28 08:35:22 <Rotti> and what does fail?
May 28 08:35:51 <Rotti> or what kind of signal do you mean? which frequency?
May 28 08:37:21 <foreverska> The coil signal varies in practice but is always a digital signal.  36 tooth timing wheel with 5 degree teeth rotating at minimum 600RPM.
May 28 08:38:35 <Rotti> so at least 36*10 hz?
May 28 08:38:59 <foreverska> yea
May 28 08:40:41 <Rotti> can you try to make C11 smaller?
May 28 08:41:18 <foreverska> How small?  C12 small?
May 28 08:42:15 <Rotti> for testing, you could leave it out
May 28 08:44:29 <foreverska> Alright.  I have to re setup the muxing.  It reset earlier.
May 28 08:45:43 <Rotti> although i guess the engineers know why they put it there...
May 28 08:45:57 <foreverska> Yea, it didn't do anything
May 28 08:48:57 <foreverska> Just to make sure I enabled my testing PRU core (it makes an ideal output) and it registered that just fine.
May 28 08:49:28 <Rotti> is the circuit on a breadboard?
May 28 08:49:36 <foreverska> yea
May 28 08:49:52 <Rotti> are all contacts okay? sometimes, i had some problems
May 28 08:50:44 <foreverska> Everything shoots fine.
May 28 08:51:02 <dogrocket> anyone using debian minimal package?  i baked my own kernel and dpkg and apt-get have file lock problems
May 28 08:51:21 <foreverska> Maybe the wheel output from the simulator isn't strong enough to switch the opto.
May 28 08:52:05 <foreverska> Seems relatively unlikely though.  They're not exactly high current devices.
May 28 08:54:46 <foreverska> Oh... it seems to be... taking the lead off the simulator output and just wiggling it on the raw 12v from the device kickstarted the PRU.
May 28 08:55:07 <foreverska> Put the lead back on the simulator and it went dead.
May 28 08:59:07 <foreverska> I'll sort this in the morning.  It's 4am I'm just now realizing.
May 28 09:07:44 <johnwalkr> yes my rev Cs or whatever these element14 ones are called came
May 28 09:11:07 <mrpackethead> http://cdn.diply.com/img/4f2c89a6-d8d9-4c6d-8f96-41e6e9d08dd7.jpg
May 28 11:12:16 <dj> guys I'm new to the xbmc thing
May 28 11:12:38 <dj> can some one plz tell me abut it
May 28 11:13:48 <dj> wat are repository
May 28 11:13:49 <dj> ?
May 28 11:32:31 <cioban> Hi
May 28 11:33:19 <cioban> I want to use a custom kernel, and a custom Pinmux configuration
May 28 11:34:14 <cioban> But I did find how I can configure pinmux without capemgr, just using the kernel...
May 28 11:34:28 <cioban> Ops I didn't find
May 28 11:35:29 <cioban> Anyone did this? Configure the pinmux for more GPIO without capmgr? I need some help.. :)
May 28 14:20:01 <thurgood> played a bit with the BBB running android last night... doesn't really seem ready for prime time, sadly :/
May 28 14:38:12 <mammbo3> So, I compiled Android 4.2.2 from the rowboat source and followed the instructions on TI's website. The not-so-funny thing is that HDMI will not display unless I boot up the system and wait ~30 seconds before plugging in the HDMI. Do you have any idea as to why that could be?
May 28 14:40:00 <mammbo3> Logcat of boot when HDMI is plugged in: http://pastebin.com/5R3sWJkf
May 28 14:41:09 <mammbo3> Oh, and this is on the beaglebone black
May 28 14:43:04 <thurgood> do you get signal out of range?
May 28 14:46:19 <mammbo3> No signal at all. I modified the da8xx-fb.c file to use 1024x768@60hz as the resolution to work with a certain touchscreen.
May 28 14:50:31 <thurgood> I was having to wait about 5 seconds or so to plug in my hdmi or I would get a signal out of range error on my monitor, but that seems to have gone away
May 28 14:53:15 <mammbo3> Yea. I don't remember having the issue when the resolution was at 720p.
May 28 14:53:15 <thurgood> I do have to pull the power completely to reboot though or I get the same problem again
May 28 14:54:41 <thurgood> I did notice that my modes file for fb0 said 1280x720p-75, even though I specified 60 in the uEnv.txt command line
May 28 14:59:08 <mammbo3> Using the TI build (3.2 kernel), I have not been able to get the uEnv.txt commands to work when trying to change the resolution. What should the syntax be?
May 28 14:59:53 <thurgood> I'm not convinced that what I specified is taking effect
May 28 15:00:10 <mammbo3> ok gotcha
May 28 15:01:48 <thurgood> I think it's still trying to get something from the edid in the monitor
May 28 15:02:24 <thurgood> I also have no audio over the hdmi
May 28 15:11:51 <mammbo3> yea, I am sort of scared that that is the case. I have tried three different HDMI to VGA adapters and they all seem to be doing the same thing.
May 28 15:12:47 <mammbo3> I then tried hooking up to an HDMI display and it is not displaying either unless plugged in 5 seconds after boot.
May 28 15:31:09 <riadNassiffe> Hello, does someone know if the BeagleBone Black or BeagleBone-xM can switch frequency in a continuous range as raspberry pi?
May 28 15:35:07 <thurgood> which frequency, pwm?
May 28 15:41:05 <riadNassiffe> I am working in a project where I need to control the frequency to save energy and avoid temperature to increase, so I need to set any integer number from 500MHz to 1GHz. Does the beagleboard can do this?
May 28 15:41:19 <Rickta59> yes
May 28 15:42:12 <riadNassiffe> any model?
May 28 15:42:23 <Rickta59> it is a linux thing
May 28 15:43:53 <riadNassiffe> Thanks, I asked because I saw that the cpufre-inf returned  available frequency steps: 300 MHz, 600 MHz, 800 MHz, 1000 MHz. But it should be some drive issue
May 28 15:44:34 <Rickta59> sorry I misunderstood your question
May 28 15:44:46 <Rickta59> you want to be able to set to say 500100 Hz?
May 28 15:45:07 <Rickta59> there are only the steps available
May 28 15:45:18 <Rickta59> 300,600,800, 1000Mhz
May 28 15:45:47 <aferm> Does anyone know if it would be possible to order the beaglebone black with custom population options? I work for a company interested in using it in an industrial setting.
May 28 15:45:54 <riadNassiffe> Yes
May 28 15:47:21 <riadNassiffe> Thanks Thurgood and Rickta59 for your help
May 28 15:48:30 <aferm> Is there somewhere I would be able to find information on how to do this?
May 28 15:50:27 <Defiant> so neo900 has been canceled?
May 28 15:50:32 <Defiant> echan
May 28 15:53:12 <myself> aferm: email circuitco
May 28 15:56:14 <thurgood> aferm:  it should automatically scale by default
May 28 16:04:29 <aferm> myself: thank you
May 28 16:32:35 <Amperture> My BBWhite seems to have a weird problem, when I plug the board into a USB port on my machine, it won't always start up the web server, the one that's supposed to be at 192.168.7.2
May 28 16:35:13 <agmlego> But it otherwise boots normally and starts all other expected services?
May 28 16:35:34 <agmlego> What do the logs have to say?
May 28 16:47:19 <Amperture> Actually I can't even get into it
May 28 16:47:31 <Amperture> I try to ssh in, and the connection gets refused.
May 28 16:48:05 <Amperture> But there's some sort of activity going on, certainly, since I can access the "Getting Started" data.
May 28 16:49:54 <agmlego> Via what, the mass-storage gadget?
May 28 16:50:12 <Amperture> I think that's it, yeah.
May 28 16:50:36 <agmlego> It sounds a lot like the bone is not actually booting up, which is probably a power issue. Try either a powered hub, or better, a DC power supply into the barrel jack.
May 28 16:51:09 <agmlego> Alternatively, watch the boot process over the serial interface and see where it goes wrong.
May 28 16:54:02 <Amperture> Hmm
May 28 16:54:15 <Amperture> I think I have reliably found a way to get it to boot outside of that
May 28 16:54:45 <_SY_> hello
May 28 16:54:55 <_SY_> is there anyone involved in Angstrom development here?
May 28 16:55:28 <_SY_> There are numerous sites with links to http://www.angstrom-distribution.org/toolchains/   which all fail as it doesn't exist 404.
May 28 16:57:03 <thurgood> if he'd stick around he might get an answer
May 28 16:57:57 <AE7IK> Maybe he found the answer he was looking for.
May 28 16:58:21 <thurgood> doubt it, he's been hit-and-running for a while now
May 28 17:00:29 <thurgood> believe he's looking for https://github.com/Angstrom-distribution ... if the logger is still active
May 28 17:01:15 <Amperture> Weird, now it's coming up with no issue.
May 28 17:01:47 <Amperture> but yeah, I got a COM port set up ready to watch the goings on over UART
May 28 17:01:56 <Amperture> next time it happens
May 28 17:02:10 <Amperture> and I've been meaning to get myself a wall wart for the thing
May 28 17:02:26 <Amperture> but no one around here carries 5V that aren't hideously expensive
May 28 17:02:32 <myself> and then it turns out that the uart is changing some grounding or power connection and making the behavior go away
May 28 17:05:10 <Amperture> Aha there we go
May 28 17:07:19 <Amperture> No, it's definitely booting the board.
May 28 17:09:12 <Amperture> http://pastebin.com/xiWFdpkY
May 28 17:09:43 <Amperture> Note the last two lines, I think they're relevant to what we're looking for, but I'm still a little wet behind the ears
May 28 17:11:03 <thurgood> are you doing a soft reset?
May 28 17:11:22 <thurgood> or pulling and reapplying power?
May 28 17:12:57 <Amperture> soft reset being running "reboot" in the terminal? No, I've done both pulled and reapplying power (in this case, USB connection), and hitting the physical button on the board.
May 28 17:14:25 <Amperture> Although, now that I know I can get into the board via UART, I'm not so concerned about the ssh and web servers not running immediately.
May 28 17:14:35 <Amperture> Although I should be
May 28 17:15:56 <thurgood> by soft I mean not completely killing the power source
May 28 17:16:31 <Amperture> I'm not 100% certain what the physical reset button does
May 28 17:16:52 <Amperture> but I have before done a yank and replace on the USB cable, which is what's supporting power right now
May 28 17:17:05 <Amperture> The issue seems indiscriminant
May 28 17:17:31 <Amperture> sometimes the web and ssh servers just won't be accessible (or won't start) despite the OS booting
May 28 17:17:46 <Amperture> Which seems to me like a network error wholesale if it's both the web and ssh servers
May 28 17:20:55 <thurgood> both are through the otg usb system, right?
May 28 17:21:13 <thurgood> or are you plugged into a router/wireless?
May 28 17:21:57 <Amperture> Yes, the only cable/wire going to my BB is the USB from my PC
May 28 17:29:02 <Amperture> Now this is really weird
May 28 17:30:14 <Amperture> During boot, I've noticed that if it boots with THIS error code, the network will be accessible
May 28 17:30:17 <Amperture> [   18.241144] systemd-udevd[85]: worker [103] failed while handling '/devices/ocp.2/47400000.usb/musb-hdrc.0.auto/gadget/net/usb0'
May 28 17:32:02 <thurgood> pretty sure that's the standard USB host, not the otg
May 28 17:32:24 <Amperture> my mistake
May 28 17:32:28 <thurgood> but I could be mistaken
May 28 17:32:51 <Amperture> but for some reason that error code came up on boot only on the times I can access the network services on the board
May 28 18:04:49 <upwardindex> When I build kernel 3.14 with rnc’s scripts my bbone boots fine, if I build the 3.8, I get nothing past “Uncompressing linux… done, booting the kernel…”
May 28 18:05:13 <upwardindex> How can i troubleshoot this?
May 28 22:08:53 <dogrocket> is there a linux program that will take image of sd card but not have to get all the free space on it?
May 28 22:10:57 <dogrocket> can DD command do it?
May 28 22:16:30 <foreverska> Yea, just tell DD exactly how much to copy
May 28 22:24:43 <vagrantc> dd would likely truncate the filesystem and result in a corrupt filesystem
May 28 22:25:43 <BobCat> dd == diskdestroyer, be careful
May 28 22:27:32 <XorA> dd to gzip/bzip2/compressor of choice
May 28 22:31:23 <vagrantc> you could grab individual partitions with partimage, but not the raw device...
May 29 01:04:21 <mjgardes> i used to generate large files filled with repeating bytes to replace the garbage in free space make compression more effective
May 29 01:05:06 <mjgardes> 0xFBFB... compresses really well with bzip2
May 29 01:18:08 <Russ> or you could just use /dev/zero....
May 29 01:18:35 <Russ> dd if=/dev/zero of=blank
May 29 01:18:37 <Russ> rm blank
May 29 01:19:50 <myself> yeah, it'll die when the device is full
May 29 01:20:03 <myself> funnier if you try to do it on a compressing filesystem without realizing it
May 29 01:20:19 <myself> you've stored 60PB of zeroes and it's still going, you're scratching your head..
May 29 01:21:28 <veremit> bs=1M count=XYZ :p
May 29 01:23:30 <ds2> just make it sparce
May 29 01:24:03 <Russ> if you are dd'ing an entire partition, it doesn't know
May 29 02:40:41 <mjgardes> yes, /dev/zero also works, but what's interesting is that 0xFB compresses slightly better than zeros in bzip2
May 29 02:54:32 <mjgardes> up to 1800000:1 approximately
**** ENDING LOGGING AT Thu May 29 03:00:00 2014