**** BEGIN LOGGING AT Tue Mar 28 03:00:04 2017
Mar 28 04:20:23 <kiran4399> ds2: there?
Mar 28 07:10:43 <l0ftyWhizZ> Hey guys! :)
Mar 28 07:46:06 <Abhishek_> I haven't registered as a mentor yet. Can an invite be sent to me?
Mar 28 07:58:28 <l0ftyWhizZ> Any BeagleBoard officials here? :)
Mar 28 11:51:27 <ravikp7> jkridner:  I did my homework on u-boot. I think I've got everything sorted out. Firstly, earlier Kconfig wasn't supported in u-boot so PATCH was used to modify the configuration manually. Now Kconfig is supported, hence configuring has become so easy. For booting beaglebone in ums mode, u-boot needs to be configured through am335x_evm_usbspl_defconfig as
Mar 28 11:51:27 <ravikp7> this config supports SPL networking and dfu for NAND. Next, the .config needs to be modified through make menuconfig to run the ums command. In menuconfig-> CLI -> Device Access Commands-> UMS , this needs to be included.  And that's it, u-boot and spl binaries can be build now with default ums mode.  And I also figured out that RNDIS vid and pid for SPL are
Mar 28 11:51:27 <ravikp7> already present in ether usb drivers, so need to specify that. Now everything right ?
Mar 28 11:58:33 <kiran4399> jkridner ds2 m_w nerdboy: I could get the SGX working on my beaglebone black..
Mar 28 12:58:27 <yugo43> m_w it's been a few days and I still haven't gotten accepted, should I just try again with a new account or wait it out a bit longer
Mar 28 12:58:34 <yugo43> I'd like to get my proposal going ASAP
Mar 28 14:26:06 <ordsen> m_w: hey have you read what I wrote yesterday about the 2nd proposal?
Mar 28 15:43:40 <pmezydlo> m_w:Hi, How to create a testbench for verilog module using Icestorm?
Mar 28 15:43:52 <pmezydlo> I want to use IceTime from Iceprog but I don't know how to visualization output
Mar 28 15:49:22 <pmezydlo> One of the projects element should be testbench files
Mar 28 16:02:17 <ds2> hmmm
Mar 28 16:05:59 <pmezydlo> ds2: hi do you have any idea
Mar 28 16:06:15 <m_w> ordsen: yes
Mar 28 16:07:10 <m_w> pmezydlo: ask on the #yosys channel
Mar 28 16:07:15 <ordsen> m_w: Do you think this a reasonable project?
Mar 28 16:07:20 <ordsen> oh
Mar 28 16:08:00 <ordsen> I found DeviceHive, they seem to have implemented a open source IoT solution, but no support for I2C, SPI, UART etc
Mar 28 16:09:13 <m_w> pmezydlo: http://stackoverflow.com/questions/35927650/is-it-possible-to-create-a-simulation-waveform-from-yosys-output
Mar 28 16:10:09 <ordsen> m_w: And.. my project is not really IoT as it does not depend on a server / connect to ethernet. So it will be different from their product and add value to the BB community
Mar 28 16:10:10 <m_w> ordsen: it sounds like an okay idea but I would want to see a bit more detail
Mar 28 16:10:38 <m_w> ordsen: you have a document of the idea yet?
Mar 28 16:11:39 <ordsen> m_w: Not yeet, I wanted to hear what you think about it first. But I can add the project proposal to the ideas list and add there more details
Mar 28 16:12:17 <ordsen> I will do that today if I can
Mar 28 16:12:27 <m_w> ordsen: okay can you summarize it again to get my brain working
Mar 28 16:13:15 <m_w> trying to figure out if there is already something similar
Mar 28 16:13:49 <m_w> or I can go back and look at the logs
Mar 28 16:14:13 <ordsen> m_w: I can summarize it
Mar 28 16:14:20 <m_w> http://logs.nslu2-linux.org/livelogs/beagle-gsoc/beagle-gsoc.20170328.txt
Mar 28 16:14:36 <m_w> but go ahead maybe new ideas will come out
Mar 28 16:16:17 <ordsen> m_w: its moving more towards a serial control application: Configuration of the serial ports happens inside the first part (C library). This exposes an interface to read from/write to serial devices.
Mar 28 16:16:40 <ordsen> On top of that is a WebServer which exposes that interface via an HTML interface. The erb UI will utilize that interface
Mar 28 16:17:36 <ordsen> so basically, the architecture remains the same like in the terminal server project. I would add functionality to control the serial devices to the web UI
Mar 28 16:18:17 <m_w> ordsen: yeah that is fine
Mar 28 16:18:29 <ordsen> and make the HTML interface accessible through the local network or locally on the BB later / as stretch goal
Mar 28 16:18:37 <ds2> pmezydlo: sorry, no.
Mar 28 16:19:05 <ordsen> m_w: Ok, I'll provide a project description and add my proposal! :)
Mar 28 16:19:27 <ordsen> m_w: Btw, did you have a look on my terminal server proposal yet
Mar 28 16:19:48 <m_w> ordsen: yes, are you just going to update that proposal?
Mar 28 16:20:01 <pmezydlo> ds2: np, We should invite Marek here,
Mar 28 16:20:14 <ordsen> m_w: uhm. I intended to write a 2nd
Mar 28 16:20:23 <ordsen> not sure yet which one I would prefer
Mar 28 16:21:00 <m_w> ordsen: they sound pretty similar in ways
Mar 28 16:21:25 <m_w> maybe seeing the to proposals might clear up the distinction
Mar 28 16:22:09 <m_w> probably just moving away from a terminal to a control application
Mar 28 16:22:15 <ordsen> m_w: Yes, they are indeed.I will point out how it is different from the terminal server.
Mar 28 16:22:29 <m_w> okay good
Mar 28 16:22:29 <ordsen> is it ok to take some parts from my existing proposal?
Mar 28 16:22:38 <m_w> sure
Mar 28 16:23:18 <ordsen> ok great.
Mar 28 16:25:20 <ordsen> m_w: I guess as the other terminal server proposal does not focus that much on the web / HTML interface stuff it won't replicate much functionality
Mar 28 16:27:51 <ordsen> m_w: Do you have any suggestions how I can improve my terminal server proposal?
Mar 28 16:29:58 <m_w> lemme see
Mar 28 16:30:26 <ordsen> http://elinux.org/BeagleBoard/GSoC/BeagleBone-based_serial_terminal_server
Mar 28 16:34:16 <m_w> ordsen: looks pretty good
Mar 28 16:36:46 <ordsen> m_w: thanks! Then I'll just write the 2nd one in a similar style
Mar 28 16:38:41 <pmezydlo> m_w: When will we start working on FPGA board?
Mar 28 16:41:32 <m_w> pmezydlo: when I find some spare time
Mar 28 16:42:06 <m_w> I check probably do the schematic this weekend
Mar 28 16:42:14 <pmezydlo> m_w: yeah sure
Mar 28 16:42:22 <m_w> will you be online then?
Mar 28 16:42:41 <pmezydlo> I will be online
Mar 28 16:42:53 <pmezydlo> I want to help you
Mar 28 16:43:20 <m_w> well you have a basic specification
Mar 28 16:44:32 <pmezydlo> I will prepare more information,
Mar 28 16:52:39 <kiran4399> m_w: I got the sgx working on beagleobne
Mar 28 16:57:47 <m_w> kiran4399: fun
Mar 28 16:58:01 <m_w> you get it to do anything interesting yet?
Mar 28 16:58:27 <kiran4399> m_w: no.. but I think I saw something interesting..
Mar 28 16:58:36 <kiran4399> m_w: http://image.inha.ac.kr/paper2/IEICE12Park.pdf
Mar 28 16:58:52 <kiran4399> this guy implemented 3 major image processing algos on an sgx540 gpu
Mar 28 16:59:12 <kiran4399> and compared their performance with a 1Ghz arm a8 processor
Mar 28 16:59:37 <kiran4399> system which is similar to the beaglebone blackj
Mar 28 17:02:47 <ds2> kiran4399: what langauge?
Mar 28 17:02:52 <ds2> GLES2? GLES3? CL?
Mar 28 17:02:57 <kiran4399> ds2: gles2
Mar 28 17:03:09 <ds2> interesting indeed
Mar 28 17:03:12 <kiran4399> ds2: check this link http://image.inha.ac.kr/paper2/IEICE12Park.pdf
Mar 28 17:05:39 <ds2> was there another link to code?
Mar 28 17:06:13 <ds2> having code to show doing soemthing like that on the bone would be the jist of the project
Mar 28 17:07:24 <kiran4399> ds2: unfortunately, I did not see any link to their code..
Mar 28 17:07:45 <kiran4399> ds2: but I could run some test programs on the sgx
Mar 28 17:08:49 <ds2> kiran4399: what is your background?
Mar 28 17:10:05 <kiran4399> ds2: I am 4th year undergraduate..  doing research in robotics.. especially in perception, vision and deep learning on computer vision
Mar 28 17:10:48 <kiran4399> ds2: https://pastebin.com/Uajf7XxP
Mar 28 17:12:24 <kiran4399> ds2: I also love hacking embedded robotics..
Mar 28 17:13:01 <ds2> kiran4399: so you are familar with the alg side of stuff?
Mar 28 17:13:16 <kiran4399> ds2: yeah.. a lot of..
Mar 28 17:13:28 <kiran4399> ds2: I am a comp science guy.. so I have an idea of alg
Mar 28 17:13:55 <ds2> kiran4399: where are you on OpenGLES?
Mar 28 17:14:21 <ds2> kiran4399: what kernel is that pastebin from?
Mar 28 17:15:30 <kiran4399> ds2: Last 2 days I've been learning some tutorials on OpenGLES 2.0..
Mar 28 17:15:46 <kiran4399> ds2: that kernel is Linux beaglebone 4.4.56-bone17 #1 Sat Mar 25 14:24:33 UTC 2017 armv7l GNU/Linux
Mar 28 17:17:14 <ds2> kiran4399: ah... somewhat new to the GLES side then? (do you know GLSL before this?)
Mar 28 17:17:48 <kiran4399> ds2: I used to program in openGL before..
Mar 28 17:17:58 <kiran4399> ds2: sort of know CUDA as well..
Mar 28 17:18:07 <ds2> kiran4399: that's newer... I have code that does the EGL init stuff.. did you see the demos for GPGPU with regular OpenGL?
Mar 28 17:18:20 <kiran4399> ds2: yeah..
Mar 28 17:18:30 <ds2> kiran4399: nice...
Mar 28 17:18:31 <kiran4399> ds2: I did sometime ago..
Mar 28 17:19:03 <ds2> kiran4399: what times are you around?
Mar 28 17:19:05 <kiran4399> ds2: I am actually working on the stereo support for black..
Mar 28 17:19:26 <kiran4399> ds2: I want to do it on the sgx530
Mar 28 17:19:36 <ds2> what do you mean by stereo support?
Mar 28 17:19:42 <kiran4399> ds2: stereo matching
Mar 28 17:19:50 <ds2> what is the output of that?
Mar 28 17:19:55 <ds2> a depth/disparity map or?
Mar 28 17:20:07 <kiran4399> ds2: exactly
Mar 28 17:20:10 <kiran4399> ds2: https://www.amazon.com/ELP-Megapixel-Camera-Module-Biometric/dp/B00VG32EC2/ref=sr_1_2?s=electronics&ie=UTF8&qid=1490468087&sr=1-2&keywords=stereo+camera
Mar 28 17:20:14 <ds2> Oooooooh nice
Mar 28 17:20:30 <kiran4399> ds2: The project is interesting one right?
Mar 28 17:20:40 <kiran4399> ds2: can it be considered for GSoC?
Mar 28 17:20:42 <ds2> is it opencv based?
Mar 28 17:21:29 <ds2> kiran4399: possibly... I suspect we need a little more mentoring resources... esp. in the area of the algs
Mar 28 17:22:02 <ds2> <--- strongly interested in openly licenseable algs written in C not using opencv
Mar 28 17:24:45 <kiran4399> ds2: it would take hell lot of time and resource to write them..
Mar 28 17:24:46 <kiran4399> :(
Mar 28 17:25:37 <kiran4399> ds2: what are your expectations from the project?
Mar 28 17:25:37 <ds2> kiran4399: there are other stuff that it can be based on...
Mar 28 17:25:43 <kiran4399> ds2: example?
Mar 28 17:26:04 <ds2> kiran4399: libccv, libpcl (more of C++)...
Mar 28 17:26:18 <ds2> libpcl has some untested stereo stuff already
Mar 28 17:26:32 <ds2> opencv is being too big of a blackbox, IMO
Mar 28 17:27:05 <kiran4399> ds2: libpcl is pointcloud library right?
Mar 28 17:27:41 <ds2> kiran4399: 1 - identifying a GSoC timeframe doable chunk. 2 - reuseable bits so others can put in their own algs (main GSoC "value" for the beagle community is to get people to leverage the GPU)
Mar 28 17:27:49 <ds2> kiran4399: yep
Mar 28 17:28:19 <ds2> kiran4399: I think those 2 should give you the foundation you would otherwise pull from opencv (correct me if I am wrong)
Mar 28 17:28:39 <kiran4399> ds2: no.. you are right..
Mar 28 17:28:47 <kiran4399> ds2: these 2 should be eneough..
Mar 28 17:29:03 <kiran4399> ds2: after all.. no body is expecting beaglebone to run deep nets..
Mar 28 17:29:08 <kiran4399> lmao!!
Mar 28 17:29:26 <ds2> another angle is (not sure if there is time) to add SGX accel to the libccv stuff
Mar 28 17:29:29 <ds2> *nod*
Mar 28 17:30:01 <kiran4399> ds2: That's awesome.. I will make time for that..
Mar 28 17:30:34 <kiran4399> ds2: is there anyother library you know of?
Mar 28 17:30:45 <kiran4399> ds2: apart from those 2?
Mar 28 17:30:46 <ds2> it isn't so much making time as "convince us it is doable"
Mar 28 17:31:15 <ds2> kiran4399: that's off the top of my head... might be others... IIRC - I couldn't make heads or tails of others so I gave up
Mar 28 17:31:24 <ds2> GSoC is relatively short
Mar 28 17:31:37 <ds2> biting off too big of a chunk has gotten a lot of students in trouble in prior years
Mar 28 17:32:10 <ds2> much rather have a well defined, doable project, and have stretch goals then a large unclear project
Mar 28 17:32:55 <ds2> there can always be improvements post GSoC
Mar 28 17:33:58 <kiran4399> ds2: then in that case.. I will consolidate a "well to do" project..
Mar 28 17:35:13 <ds2> write a proposal/app
Mar 28 17:35:25 <ds2> talk to other mentors to get their input
Mar 28 17:35:36 <kiran4399> ds2: I hope you know that I am that old kiran4399?
Mar 28 17:35:38 <kiran4399> :)
Mar 28 17:35:47 <kiran4399> last time gsoc
Mar 28 17:35:48 <ds2> kiran4399: what times are you around? this week I have a crazy schedule
Mar 28 17:36:14 <ds2> kiran4399: I don't track that...was guessing that was the case but....
Mar 28 17:36:52 <kiran4399> 9:00 AM to 11:55PM IST
Mar 28 17:36:55 <kiran4399> ds2:: ^^
Mar 28 17:37:00 <kiran4399> ds2: what times you are free?
Mar 28 17:37:16 <ds2> let me translate IST
Mar 28 17:37:53 <ds2> I am on and off at random times
Mar 28 17:38:18 <ds2> I can try to make mornings work but this THU/FRI I have other things
Mar 28 17:38:37 <kiran4399> are you free this time..
Mar 28 17:39:02 <ds2> in the future I can try to be around at this time but this week is just crazy on my end :(
Mar 28 17:40:14 <kiran4399> ds2: are you sure there is only libccv?
Mar 28 17:40:21 <kiran4399> ds2: is there anything better?
Mar 28 17:40:57 <ds2> kiran4399: there are others... libccv was the most straightforward for me (a non alg guy)
Mar 28 17:41:07 <ds2> I have used libccv and it just largely works
Mar 28 17:41:22 <ds2> might be better stuff... opencv (from source) is a mess
Mar 28 17:41:33 <ds2> libpcl is less of a mess
Mar 28 17:41:56 <ds2> libccv's face detect was pretty easy to do...just a tad bit slow on the A8
Mar 28 17:42:56 <ds2> kiran4399: I do watch this channel... so you leave a URL with my nick on it...i.e. "ds2: http://foo.bar/", I'll probally eventually see it
Mar 28 17:46:54 <kiran4399> ds2: this also looks good..
Mar 28 17:46:57 <kiran4399> https://www.edwardrosten.com/cvd/
Mar 28 17:48:02 <thetransformerr> hi everybody!!!!!!
Mar 28 17:48:27 <ds2> kiran4399: heh... C++
Mar 28 17:48:44 <kiran4399> ds2: c++ no good?
Mar 28 17:49:11 <thetransformerr> nerdboy: I hope you remember me despite your busy schedule, i am working on proposal for sonic anemometer
Mar 28 17:49:31 <ds2> let me try compiling it later
Mar 28 17:50:07 <thetransformerr> nerdboy:I today found out a paper focused on tilt correction algorithms, what do you think, is it useful for us...
Mar 28 17:50:11 <ds2> kiran4399: In my opinion, C++ is a pain - often it is picky on compiler version... i.e. works with gccXYZ but not gccXYZ+1
Mar 28 17:50:33 <ds2> kiran4399: I am flexible and can be convinced
Mar 28 17:51:52 <thetransformerr> hi ds2:, would you like to provide your review and quote on my proposal....please!!!! :)
Mar 28 17:51:54 <thetransformerr> http://elinux.org/GSoC2017_sonic_anemometer_proposal
Mar 28 17:52:31 <thetransformerr> nerdboy: please reply whenever you read above message...
Mar 28 17:54:04 <thetransformerr> kiran4399: hi buddy!!! I can feel you are doing something fancy...can you provide me any idea what you are intending to do..:)
Mar 28 17:54:35 <kiran4399> thetransformer: I am trying to port a small cv library on the beaglebone black..
Mar 28 17:54:52 <kiran4399> also by using the sgx gpu onboard.
Mar 28 17:56:24 <thetransformerr> okk, and cv here is computer vision..
Mar 28 18:01:31 <ds2> thetransformerr: do you have a PRU daq? and are you familiar with its limitations? I think you need 2 channels so the sample rate is not going to be 40Msps
Mar 28 18:04:02 <nerdboy> there should be a prudaq branch in last year's repo
Mar 28 18:05:23 <ds2> nerdboy: how fast does it need to be?
Mar 28 18:05:40 <ds2> I see no numbers but I assume you discussed it?
Mar 28 18:07:58 <ravikp7> jkridner: I've updated my proposal http://elinux.org/BeagleBoard/GSoC/node-beaglebone-boot to remove usage of u-boot patch and added info for new configuration for u-boot for default ums mode, please review it.
Mar 28 18:16:36 <nerdboy> last year was a bit lacking in communication...
Mar 28 18:17:01 <thetransformerr> ds2: I don't have pru daq currently but I have read most of the available wiki on that device
Mar 28 18:17:24 <nerdboy> he made the switch to prudaq on his own after fighting with the TI board
Mar 28 18:18:15 <nerdboy> seems compatible with the murata sensors but i have no idea about performance/data quality
Mar 28 18:18:42 <thetransformerr> ds2:as you say we won't require 40msps, but I can't quote exact number by now, as I haven't that much data or experience, you can help about it
Mar 28 18:20:50 <thetransformerr> nerdboy: I have seen such project like ours implemented with murata,
Mar 28 18:20:54 <thetransformerr> https://soldernerd.com/2017/02/06/ultrasonic-anemometer-part-30-downsized-hardware/
Mar 28 18:21:15 <kiran4399> ds2: you know what?? libcvd has opengl es2 support..
Mar 28 18:21:47 <kiran4399> ds2: I am compiling it on beaglebone... looks like it is succesfuuly compiling with GPU support!!
Mar 28 18:25:51 <thetransformerr> he used multi comp at last, but used murata in between his progress
Mar 28 18:29:43 <nerdboy> just because a couple of us have muratas doean't mean you have to use them
Mar 28 18:30:32 <nerdboy> the data signal is kinda important...
Mar 28 18:33:09 <thetransformerr> yeah nerdboy: , but either we can test available range or learn from previous experiments just for intialisation
Mar 28 18:33:51 <nerdboy> smaller is actually better as long as the data is good
Mar 28 18:36:21 <thetransformerr> and data can be tested either using DAQ or waveform monitor???
Mar 28 18:36:36 <nerdboy> the wiki page is in decent shape, what can you say about sensor specs?
Mar 28 18:36:55 <nerdboy> gotta ask for something...
Mar 28 18:37:55 <ServerProcessor> m_w : I'm planning to integrate a text editor on the interface and I'm midway creating mockups for the same to get your reviews. Do you think its a good idea to integrate a text editor ?
Mar 28 18:38:07 <thetransformerr> actually I didn't focused on that part, but as now you and ds2 pointed out , I would come up with details till tomm
Mar 28 18:38:27 <m_w> ServerProcessor: text editor?
Mar 28 18:39:10 <ServerProcessor> m_w : In a way user can modify the C libraries if they wish to.
Mar 28 18:40:03 <m_w> I am confused
Mar 28 18:49:24 <ServerProcessor> m_w : I was thinking of uploading scripts through the interface or modify it.
Mar 28 18:51:07 <ServerProcessor> To improve user experience.
Mar 28 19:00:52 <m_w> C is not a scripting language
Mar 28 19:04:43 <ServerProcessor> m_w : Sorry,I meant C programs.
Mar 28 19:06:07 <amil101> Hi!
Mar 28 19:06:14 <amil101> I'am creating the proposal for the Google Summer of Code and the proyect I have in mind is to create a opening system for Makespaces, Hackspaces, Fablabs, etc. The problem is that when the user of this of comunity don't pay, he must not enter in the space. So my solution is to create opening system with NFC, that when the user pass a card, the beagleBone check in the database if the user is in the space. I think that this can be 
Mar 28 19:06:22 <amil101> What do you think?
Mar 28 19:07:31 <amil101> Sorry about sending the email in the #beagle chat, I didn't know this group exists
Mar 28 19:22:39 <jkridner> amil101: doesn't sound like broad support for the beagleboard.org community, but more of an instance project. do you have a large number of people asking for this already?
Mar 28 19:23:06 <amil101> nope
Mar 28 19:23:30 <jkridner> I don't think there is a lot of demand for such a thing as most spaces can manage to create their own (and have done so).
Mar 28 19:23:50 <jkridner> there's some room for collaboration on it, but I think you have to take a bigger picture in terms of what you are creating.
Mar 28 19:23:54 <amil101> thanks, that was one of my doubt of the proyect
Mar 28 19:24:12 <amil101> I will think about it
Mar 28 19:24:31 <jkridner> what libraries are needed for such a project? which ones already exist? what needs to be created?
Mar 28 19:37:07 <ravikp7> jkridner: am I right about the u-boot configuration now which I described in earlier messages ?
Mar 28 19:46:40 <ordsen> m_w: Almost forgot: I got my BB and made the UART experiment we talked about last week..
Mar 28 19:48:29 <ordsen> m_w: while spamming the UART port not a single package was lost, I will extend the test a bit but for the moment I think that the I/O buffering could be implemented without modifying the serial driver
Mar 28 19:48:48 <jkridner> ravikp7: it sounded right.
Mar 28 19:49:00 <jkridner> ravikp7: you are welcome to send me some binaries to test.
Mar 28 19:50:51 <m_w> ordsen: good
Mar 28 19:54:38 <ravikp7> jkridner: Thanks for your reply, I was awake for this, time to sleep now for me , 1:20 am here. I'll send you binaries later today.
Mar 28 19:54:52 <jkridner> k
Mar 28 21:40:49 <ds2> thetransformerr: you should be able to estimate the required sample rates
**** ENDING LOGGING AT Wed Mar 29 03:00:03 2017