**** BEGIN LOGGING AT Fri Oct 14 02:59:58 2016 Oct 14 03:12:08 Set_: "Time Series Analysis by State Space Methods" and "Understanding Digital Signal Processing" Oct 14 03:12:47 Set_: oh.. and "The Left Hand of Darkness" Oct 14 03:19:10 Horrowitz and Hill: The Art of Electronics. Oct 14 03:19:29 oh sorry .. I was thinking of recommended texts .. lol Oct 14 03:26:31 Oh. Oct 14 03:27:06 KotH: good reading there :D Oct 14 03:27:08 I am reading a book, "Embedded Linux Primer." It is my first interesting book. Oct 14 03:27:46 Man...I got to know what the hell it is that I keep complaining about. Oct 14 03:29:04 I got this funny "course" from computer.org and this fellow speaks funny. I cannot understand what it is that he is trying to convey. Oct 14 03:29:59 lol Oct 14 03:30:01 ...I think he just wants to yell about embedded projects. Oct 14 03:30:25 computer.org is pretty shitty for anything but scientific conferences... and even these are often pretty low brow Oct 14 03:31:05 Boo! I know now. I might get a cert. from them, i.e. when I get rich and wealthy. Oct 14 03:31:27 He reminds me of this professor that kept yelling at math. Oct 14 03:32:24 FOUR TIMES THIS "N" NUMBER MINUS 14 OVER X(X -4)/4 IS MATH! Oct 14 03:38:23 So, what is going on in the world of the BBB? Oct 14 03:43:56 I got the Bluetooth on that BBGW up and running. No file sharing...yet! Oct 14 03:44:56 The fellow said they have an issue. There cannot be a bluetooth and WiFi working at the same time. Oct 14 05:31:13 KotH: I just finished Left Hand recently Oct 14 05:31:18 Cool Oct 14 05:35:33 ericxdu: i'm re-reading it :) Oct 14 05:36:57 This was my first time Oct 14 07:08:37 I don't have access to it right now... out earlier I was trying to use 0launch and my bbb seemed to be miscalculating a hash and rejecting the archive Oct 14 07:09:38 Its an odd problem. I'll try it again soon. The way things are going, I do *not* want Zeroinstall to fail on any of my devices Oct 14 07:47:38 Anyone having success with getting SPI transactions running on the BBB? Oct 14 07:48:14 all the ioctl()s return success, but nothing comes out the pins Oct 14 07:50:39 hello Oct 14 07:50:50 anybody here？ Oct 14 07:55:08 morning Oct 14 07:57:32 hiii Oct 14 08:02:45 hello Oct 14 08:03:29 i have a question, would somebody help me? Oct 14 08:06:59 is anybody here? Oct 14 08:23:21 hello Oct 14 08:23:24 hello Oct 14 08:23:26 hello Oct 14 08:25:54 hi Oct 14 08:29:39 hi Oct 14 08:30:37 where could I buy the beaglebone MSTP Cape? Oct 14 08:31:37 Is anyone know it? Oct 14 08:32:11 where could I buy the beaglebone MSTP Cape?Is anyone know it? Oct 14 09:32:07 dsp's work now thank u x15 Oct 14 09:32:27 all i had to do was downgrade the kernel, mess around with dkms for a few days and pray to an ancient evil Oct 14 09:33:58 I now lack a soul but at least i my float calculations have a 2,5 to 3x speedup Oct 14 11:17:31 Has anybody interfaced 4 HC-SR04 sensors to beaglebone black? Oct 14 12:01:52 hai where i get the kernel 3.8.13-bone47 for my Beaglebone black for compilation purpose Oct 14 13:38:19 hello, trying out the BBB with a recent debian console image... and a rotary encoder that i'd like to get data through interrupts. Oct 14 13:38:42 i see some tutorials online by drew fustini on eqep, and i see the beagle pru page as well. Oct 14 13:39:49 and i'm trying libmraa and python. but still not entirely sure which way to go. Oct 14 13:39:57 hi jkridner Oct 14 13:42:31 repeating for jkridner: hello, trying out the BBB with a recent debian console image... and a rotary encoder that i'd like to get data through interrupts. Oct 14 15:07:39 rajesh: I have a qep example in https://github.com/mvduin/py-uio Oct 14 15:07:52 I think there's also a kernel driver but it's quite a bit more limited iirc Oct 14 15:14:39 for my original application I personally opted to use gpio instead though since the rotation was very slow (turned manually) and I wanted an irq on every step Oct 14 15:18:19 zmatt: thanks for that, i'll have a look. my application currently does a 1ms poll, which i am hoping to replace with the interrupts thro' eqep, or the pru. Oct 14 15:18:37 all this is new to me, so maybe what i'm saying/intending isn't entirely accurate. Oct 14 15:18:53 but the idea is to move away from polling, to interrupt-driven. Oct 14 15:21:39 well eqep is of course made for the task, it can do position tracking, velocity measurement, and includes a watchdog for motor stall detection Oct 14 15:23:06 but since I just needed raw +1/-1 events (which seems relatively difficult with eqep without polling) and the signal was very slow anyway I opted for gpio instead Oct 14 15:23:23 without polling though Oct 14 15:23:52 that makes sense. i too only need the steps, and don't have a need for the other measurements. Oct 14 15:24:21 if you're currently usin 1ms polling with gpio that also suggests you have a very slow signal? Oct 14 15:24:50 what do you mean by 'slow' signal? Oct 14 15:25:14 relatively lot of time between steps Oct 14 15:25:42 it's possible. here's more details about what i'm trying to do... Oct 14 15:26:32 we're trying to track animal movement on a treadmill. a rotary encoder is attached to a wheel. we'd like to know when it moves, and for how long, so timestamping the steps. Oct 14 15:26:56 someone's written arduino code that does it every 1ms and i'm trying to move to the BBB and python. Oct 14 15:27:30 you don't need to know how fast it moves? Oct 14 15:27:32 i haven't written code yet, but looking at different ways to do it. Oct 14 15:27:54 no, velocity hasn't been discussed yet, just movement. Oct 14 15:28:25 but you bring up a good question i should ask the enduser. Oct 14 15:28:40 with gpio the limiting factor would probably be that above a certain velocity you'd lose the ability to figure out the speed and direction of the rotation Oct 14 15:29:22 gpio could also miss events if the process or the system briefly hangs for whatever reason Oct 14 15:29:37 correct. Oct 14 15:29:57 precisely why i want to try eqep/pru Oct 14 15:30:01 qep would have the benefit of tracking movement in hardware, so you can simply see whether and how much the animal has moved since the last observation Oct 14 15:30:14 because i have data coming in from other channels that need processing as well. Oct 14 15:31:56 also once there's movement, if you don't care about the details you can simply arm the watchdog to get notified when the animal stops Oct 14 15:33:28 do you know of any resources/docs that discuss how to get started with eqep and the bbb? Oct 14 15:33:36 something recent... Oct 14 15:35:34 I just pointed to my py-uio repo which includes a qep demo Oct 14 15:36:01 yes, i see that. but did you follow any resources/docs as you worked on it? Oct 14 15:36:06 anything in particular? Oct 14 15:36:12 the am335x trm Oct 14 15:36:28 ok Oct 14 15:37:19 and C++ code I had previously already written for it Oct 14 15:38:56 I'm afraid btw that py-uio probably isn't going to see much activity from me... it was an experiment to make peripherals easier to use by people, hence also python as choice of language, but the more I used it the less I like python Oct 14 15:41:06 doesn't matter to me if it's python or c/c++, i'm more concerned about lts. which is also why i decided to try libmraa as being backed by intel at least gives a sense that there will be some support. Oct 14 15:43:28 never heard of it, but if you're going to use gpio for this then some complicated framework sounds like it will only decrease the maintainability Oct 14 15:43:36 especially since the decoding is so trivial Oct 14 15:44:33 i agree. fyi - https://github.com/intel-iot-devkit/mraa hadn't heard of it either until a couple days back and furious online searches and referencing. Oct 14 15:45:26 it might be relevant if you care about portability, but I generally prefer to be able to actually use the unique features available on an embedded system Oct 14 15:45:42 i also liked the possibility it gives of trying different boards with little/no application code changes. Oct 14 15:46:09 I wonder how that turns out in practice Oct 14 15:46:19 probably not pretty Oct 14 15:46:39 and whether it's worth the loss of functionality and understandability usually involved with abstraction layers Oct 14 15:46:56 tradeoffs Oct 14 15:47:24 zmatt: why didn't use just use the pybbio module instead of writing your own qep module? Oct 14 15:47:53 I'm not aware of any other uio wrappers, but I haven't searched intensively Oct 14 15:48:23 ok Oct 14 15:48:29 qep was simply a relatively easy but non-trivial example of how to use uio Oct 14 15:48:33 and useful Oct 14 15:49:10 my impression of bbio is that they're crappily written and largely pointless wrappers around kernel interfaces (typically sysfs) Oct 14 15:49:29 but maybe I remember wrong Oct 14 15:49:50 have a pointer to what you found for qep? Oct 14 15:50:07 that was my impression as well... but just instinctual.. Oct 14 15:51:17 e.g. for gpio it opens the applicable sysfs file, reads/writes it, and closes it again... instead of leaving it open and caching the file descriptor Oct 14 15:51:39 (I also see no reason why you'd do that in a python extension written in C rather than in pure pyton) Oct 14 15:53:50 you don't need to... Oct 14 15:54:01 sounds like mraa isn't particular uptodate: https://github.com/intel-iot-devkit/mraa/blob/master/docs/beaglebone.md Oct 14 15:54:08 still assumes kernel 3.8 Oct 14 15:54:23 *particularly Oct 14 15:55:27 it's not Oct 14 15:55:40 their focus would clearly be on intel platforms... Oct 14 15:57:03 on a side note, if i were to connect the power supply for a rotary encode to the bbb, what pin would i connect it to - 5 / 6 / 7 / 8? not sure if it should be connected to 5V raw or 5V... Oct 14 15:57:39 3.3V Oct 14 15:57:56 the beaglebone I/Os are absolutely not 5V tolerant Oct 14 15:58:28 so when do you use those pins? Oct 14 15:59:34 as general power supply for external hardware, or in case of pins 5-6 to power the beaglebone itself Oct 14 16:00:16 right i want to supply 5V to the rotary encoder. Oct 14 16:01:28 I'm not used to rotary encoders needing power? typically they have ground and two data pins that they pull to ground and require external pull-up? but I suppose it might vary Oct 14 16:02:30 this is what i'm looking at - http://bit.ly/2dPMJc4 Oct 14 16:02:37 note that pins 5-6 connect directly to the 5v dc barrel jack, which means that there's no 5v there e.g. if you're powering the beaglebone via usb Oct 14 16:02:56 am powering over the dc barrel jack Oct 14 16:03:23 i need to read up on basic electronics... Oct 14 16:03:30 another difference is that sys_5v delivers no power when the beaglebone is switched off Oct 14 16:03:58 which is fine. sys_5v = pins 7, 8? Oct 14 16:04:21 or 5-6 Oct 14 16:05:17 https://goo.gl/Jkcg0w my pins spreadsheet... the P9 and P8 tabs are useful as quick reference :) Oct 14 16:05:46 sys_5v is labeled vsys there since it's not strictly speaking always 5v under all circumstances (e.g. when battery powered it'll drop down to the battery voltage) Oct 14 16:06:01 normally it'll be 5v though when the beaglebone is on Oct 14 16:06:27 ok so your encoder has 5v push-pull outputs... so you'll need some kind of buffer to interface it to the beaglebone Oct 14 16:07:48 zmatt: thanks very much for your time and help here. much appreciated! Oct 14 16:08:10 you're welcome, and good luck :) Oct 14 16:08:21 I'm afk for a bit Oct 14 16:08:23 i didn't expect to have to know a lot more than - connect device to pin X, setup/configure kernel modules at most, and write application software. Oct 14 16:08:32 heh Oct 14 16:08:36 but seems like that was being very hopeful. Oct 14 16:09:07 yeah sorry, dabbling in embedded systems will typically require hardware knowledge sooner or later Oct 14 16:09:46 yeah, and mine's atrophied after almost 20 years of not using/exercising it... Oct 14 16:10:40 it is critical for the beaglebone (and most other hardware) that the voltage applied to its I/Os does not significantly exceed its I/O supply voltage (3.3V, or 1.8V for the analog inputs) Oct 14 16:11:43 this also means you may not apply any voltage at all when the beaglebone's I/O supply rail isn't powered up (yet/anymore) Oct 14 16:13:56 i think i'm following, but not sure, heh. confusion because i'm reading 5V out on the pin diagram, then 5V supply for the encoder, and wondering why i can't just connect the two. :) Oct 14 16:14:16 you can, but the outputs of the encoder can't be connected directly to the beaglebone Oct 14 16:14:28 ahh, you meant the outputs! ok. Oct 14 16:14:39 unless you like crispy fried processor I/O pins Oct 14 16:15:18 ok, got it. Oct 14 16:15:45 i was stuck on the power supply, and didn't catch that you meant the encoder output voltages. Oct 14 16:15:50 ah ok Oct 14 16:16:09 yeah vsys is fine for powering it, but it has 5v push-pull outputs apparently Oct 14 16:16:40 thank you for noticing that, i did not pay attention. Oct 14 16:18:16 and you can't even get away with using a voltage divider since vsys powers up way before the 3.3V supply does Oct 14 16:18:56 https://lh3.googleusercontent.com/-aWEH_7JAEbw/VUBn8aVmaoI/AAAAAAAAACE/5RFxipIqG_E/s1600/dc.png (clean power-up and -down, measured on various pins) Oct 14 16:23:59 the simplest solution is one similar to the one employed on the beaglebone for its serial console header (which is 5V tolerant even if the beaglebone is powered off): run the signals through a 5v-tolerant buffer powered from the 3.3V Oct 14 16:25:16 (and consider putting series resistors on the buffer outputs, 50 ohm or so) Oct 14 16:25:20 and what do you mean by buffer... like, a resistor? Oct 14 16:26:39 https://en.wikipedia.org/wiki/Digital_buffer Oct 14 16:26:39 [WIKIPEDIA] Digital buffer | "A digital buffer is a special electronic circuit that is used to isolate the input from the output, preventing the impedance of one circuit from altering the impedance of another. The output drive capability is generally much higher than their input signal requirements and is used to drive high-current..." Oct 14 16:26:43 http://www.ti.com/lsds/ti/logic/buffer-driver-transceiver-overview.page Oct 14 16:26:58 * zmatt wants to kick that annoying bot Oct 14 16:28:17 (note that inverting or non-inverting buffer doesn't really matter for this particular application) Oct 14 16:29:51 voltage translator would also be an appropriate keyword to look for (there's much overlap between the two categories) Oct 14 16:30:45 I'm not actually a hardware guy either so I have no idea which components (out of the ocean of choices) are common ones to use Oct 14 16:30:48 that term makes sense.. Oct 14 16:31:18 anyway, I was heading out :P will be back later probably Oct 14 16:31:22 the person who did this on the arduino also had a L293NE H-bridge on the board... Oct 14 16:31:45 i wonder if this was one of the reasons for its use... Oct 14 16:31:51 ok, thanks again zmatt Oct 14 16:33:09 logic level converter Oct 14 17:17:06 I only recently realized Qupzilla is based on Webkit not Firefox Oct 14 17:17:22 Its a decent replacement for Midori Oct 14 17:17:35 Still wonder why Midori is not in the repo Oct 14 18:59:45 I have a question about where to get Linux header files Oct 14 19:01:25 I've seen some old instructions referencing rcn-ee.net, but they aren't working Oct 14 19:03:19 Dricker: http://repos.rcn-ee.com/ ? Oct 14 19:03:20 Specifically running 3.8.13-bone50, looking for Linux headers. Oct 14 19:04:18 Will try that Rajesh, thanks Oct 14 19:08:37 Rajesh, I am looking for a .deb for that kernel that will install header files.....having difficulty finding it Oct 14 19:13:23 are you having difficulty finding that specific package? apt-cache search linux-headers-`uname -r` ? Oct 14 19:13:25 I can find the folder on rcn-ee.net, but it is a .diff instead of a .deb Oct 14 19:14:20 when I execute the apt-cache, all versions are much older, 3.2.0 Oct 14 19:18:20 Dricker: looking at the rcn repo, i see 3.8.13-bone headers from version 65 through 80 Oct 14 19:18:45 Are you running the image the beaglebone shipped with? Oct 14 19:19:52 What address Rajesh? Oct 14 19:20:18 rajesh: why not use 3.8.13-bone80 ? Oct 14 19:20:39 afaik the main differences should be bugfixes and such Oct 14 19:21:07 @Math, no, it came with angstrom, earlier bbb, version A5A. Oct 14 19:21:24 zmatt: i don't need it. Dricker seems to want a specific header package version. i'm assuming he's aware of what he's doing... Oct 14 19:21:33 oh sorry I meant Dricker Oct 14 19:22:09 Rajesh, 80 would be fine, I can upgrade kernel, just having issues getting kernel headers installed Oct 14 19:23:05 Dricker: what issues? Oct 14 19:23:06 Dricker: then include [deb-src] in your sources.list and get it using apt-get? Oct 14 19:23:18 there's no need for a deb-src to install headers Oct 14 19:23:39 sudo apt-get update && sudo apt-get install linux-headers-3.8.13-bone80 Oct 14 19:23:44 true. i take it back. Oct 14 19:23:45 heh Oct 14 19:23:55 thanks for correcting zmatt Oct 14 19:26:05 I run apt-get update, but it can't find any Linux headers in the 3.8.13 range, just much older Oct 14 19:26:31 huh, what repositories are listed in your /etc/apt/sources.list ? Oct 14 19:26:32 Apt-cache search only brings up 3.2.0 and earlier Oct 14 19:26:36 what repos do you have in sources.list? Oct 14 19:26:42 nvm Oct 14 19:26:44 hehe Oct 14 19:27:04 You should check your repositories list btw Oct 14 19:27:17 *shot* Oct 14 19:27:26 :P Oct 14 19:28:31 Sorry, relatively new to some of this Oct 14 19:29:21 Debian.beagleboard.org/packages Oct 14 19:29:35 Security.debian.org Oct 14 19:29:58 can you just paste the contents of the file via e.g. pastebin.com ? Oct 14 19:30:26 I'll try Oct 14 19:30:57 also, debian.beagleboard.org ? I don't think I've ever seen that repository Oct 14 19:31:19 it looks ancient Oct 14 19:31:44 Probably is Oct 14 19:31:59 Any specific ones I should add? Oct 14 19:32:10 deb http://repos.rcn-ee.com/debian/ wheezy mani Oct 14 19:32:15 *main sorry Oct 14 19:34:41 Failed to fetch, 404 Oct 14 19:35:07 Get a few other errors saying no public key available Oct 14 19:36:18 the no public key available is normal, you need to apt-get install rcn-ee-archive-keyring Oct 14 19:36:31 but that suggests it did fetch the package lists... what did you get a 404 on? Oct 14 19:36:45 Run-ee Oct 14 19:36:53 Rcn Oct 14 19:37:32 Getting keyring Oct 14 19:37:43 it most definitely does exist: you can open http://repos.rcn-ee.com/debian/dists/wheezy/ in a browser and see "main" is definitely there Oct 14 19:38:57 Wheezy/main/binary-armel/packages had the 404 Oct 14 19:39:08 o.o; Oct 14 19:39:30 this is getting an exciting experience in archeology Oct 14 19:41:26 you have a debian armel (softfloat) installation instead of armhf (hardfloat) ... Oct 14 19:42:35 Hey! It got a few errors after installing keyring and then trying update, but doing a search on Linux-header, I have a ton and the one I am after! Oct 14 19:43:18 ah, your system is trying to fetch both armhf and armel package lists? Oct 14 19:43:33 I may need to get smarter on Linux so I don't have to pester others like this Oct 14 19:44:02 I guess so, where is that set? Oct 14 19:44:03 I'm not sure I want to know how you got into that situation, but you can suppress it for specific sources by adding an [arch=armhf] qualifier, like so: Oct 14 19:44:15 deb [arch=armhf] http://repos.rcn-ee.com/debian/ wheezy main Oct 14 19:44:49 That is what I left out in source.list, I saw other links had that Oct 14 19:44:54 uhh, not sure, adding architectures to a system or removing them isn't exactly something I do frequently Oct 14 19:45:25 Is it best to leave all these header files, or just the current version Oct 14 19:46:11 I've never any kernel headers package myself... I think they're needed for dkms in which case you generally want a headers package for every kernel you have installed Oct 14 19:47:47 Ill do some checking. Thanks zMatt, I owe you! Been trying to figure this out for over 7 hrs Oct 14 19:47:53 note btw that armhf and armel are incompatible ABIs, so an armhf application needs armhf versions of shared libraries and ditto for armel Oct 14 19:48:01 so you rarely want both on one system Oct 14 19:48:39 Makes sense Oct 14 19:49:13 although with multiarch (does wheezy have proper multiarch yet? I really don't remember) they should be kept in separate dirs (e.g. /usr/lib/arm-linux-gnueabihf) hence not conflict Oct 14 19:49:45 so in that case I imagine there might be some use for it if you have an armel executable you can't recompile to armhf (e.g. if you don't have the source code) Oct 14 19:49:59 Might wipe my BBB clean and start without all the extra files Oct 14 19:51:13 both wheezy and kernel 3.8 belong in a museum imo, but maybe I'm just being neophilic here ;) Oct 14 19:52:03 the "current" (read: not-very-old) versions are jessie and kernel 4.4 Oct 14 19:52:07 No doubt, most examples for BBB I find are using 3.8 Oct 14 19:52:16 I personally run stretch and kernel 4.8 Oct 14 19:52:32 yeah there's a definite shortage of uptodate documentation Oct 14 19:53:27 I need to get comfortable first, then might try moving to newer kernels once I get the hang of it Oct 14 19:54:24 fair enough, although there's a risk that the stuff you've then become comfortable with has all changed Oct 14 19:55:10 My issue is things like that source.list.......had no clue it was there or where apt-get update was getting its addresses to search Oct 14 19:56:20 fair enough, although for figuring out general linux stuff I'd say you're better off with a recent system than an old one Oct 14 19:56:20 No doubt on the changing part. I just need to get smarter on workings of Linux....... Oct 14 19:56:54 the only benefit I see for kernel 3.8 is the amount of stuff written for it (not necessarily all good stuff though) Oct 14 19:57:28 True. Well, off to play with my new header files, thanks again, glad there are folks out there to help me get going Oct 14 19:57:59 Dricker: out of curiosity, what do you need those headers for anyway? Oct 14 19:59:35 since I think the linux-headers package is normally only used for compiling kernel modules out-of-tree, which seems like an odd thing to be doing if you're new to linux Oct 14 20:00:45 ZMatt, following a tutorial on writing drivers for the BBB Oct 14 20:00:59 oh ok, so you are compiling kernel modules out of tree Oct 14 20:01:21 ehm Oct 14 20:01:40 why? :) Oct 14 20:01:47 Should I be concerned? :) Oct 14 20:02:18 well yes, since the linux kernel is a very fast-paced and dynamic environment which doesn't care about maintaining compatibility of kernel interfaces Oct 14 20:02:30 so learning to write drivers for an ancient kernel seems quite pointless Oct 14 20:03:06 But maybe helps me understand how it works? Oct 14 20:03:48 it seems like an odd way to do so, but to each their own I guess Oct 14 20:03:55 Maybe I need to search harder for a tutorial using a newer kernel Oct 14 20:04:47 although I occasionally do write kernel stuff, and frequently consult the kernel sources if I need to know the fine details of how some particular functionality works, I would certainly not consider it an easy piece of documentation Oct 14 20:05:35 Gotcha, what beaglebone do you work with? Oct 14 20:06:42 Are there any builds for bbb that utilize 4.8 kernel? Oct 14 20:06:52 black rev c mostly (and I have one a5a with jtag connector), although the differences aren't really that relevant in practice Oct 14 20:07:35 you can simply install a 4.8 kernel on a standard jessie release Oct 14 20:08:15 I started with a jessie console install and then first upgraded to stretch Oct 14 20:09:00 latest images can be found here btw: http://elinux.org/Beagleboard:BeagleBoneBlack_Debian Oct 14 20:09:54 or, if you prefer to stick wheezy, you can find the last release (2015-06-15) here: http://beagleboard.org/latest-images Oct 14 20:09:59 *stick with Oct 14 20:10:19 Thanks! I'll see what I can tear up Oct 14 20:10:24 :) Oct 14 20:12:06 probably you can even ancient kernels on recent debians and vice versa (up to some limits anyway) Oct 14 20:12:20 *even run ... argh I keep typing ahead of my thoughts Oct 14 20:12:24 I wonder why when I upgraded to bone50, source.list wasn't updated to include that rcn address? Oct 14 20:12:48 I have no idea, I didn't even know that repo of yours existed :) Oct 14 23:39:13 hi from greece. Oct 15 02:36:19 JetForMe bb Oct 15 02:36:50 ? **** ENDING LOGGING AT Sat Oct 15 02:59:58 2016