**** BEGIN LOGGING AT Fri Mar 25 02:59:59 2016 Mar 25 03:16:07 m_w: regarding eeprom.. Mar 25 03:16:14 I think you need them for building a new image.. Mar 25 03:16:14 yes Mar 25 03:16:25 why? Mar 25 03:16:53 explain how it works Mar 25 03:17:21 on the bbb eeprom were used on the capes to let the kernel know what hardware it was Mar 25 03:17:54 but the blue is now the based board and the hardware is known Mar 25 03:18:22 hence all of the feature on the board can be built into the toplevel dts made for the blue Mar 25 03:18:49 features Mar 25 03:20:29 s/based/base Mar 25 03:21:01 is there a toplevel specification for the beaglebone blue available? Mar 25 03:24:44 I have only seen pictures of unpopulated PCBs Mar 25 03:28:18 I have also seen the strason cape: http://www.strawsondesign.com/#!board-specs Mar 25 03:29:03 "patent pending" seriously? Mar 25 03:35:15 m_w: there? Mar 25 03:35:18 m_w: see this Mar 25 03:35:20 https://github.com/beagleboard/image-builder Mar 25 03:35:45 okay Mar 25 03:41:28 okay so this is an EEPROM on the base board Mar 25 03:44:11 m_w: yeah. Mar 25 03:48:34 so you have to write a unique signature into the EEPROM and it is used in U-Boot to initialize things Mar 25 03:49:30 m_w: I guess it is already there right? Mar 25 03:49:58 I have no idea how far the board support is on the blue Mar 25 03:50:21 No.. I mean to say the EEPROM signature is already there right?? Mar 25 03:50:28 on the image_builder git repositiry.. Mar 25 03:50:57 it appears so Mar 25 03:51:51 so you wont have to deal with that directly Mar 25 03:52:35 m_w: do you know how many pwms are supported by each pru? Mar 25 03:53:19 12 I think Mar 25 03:53:34 but depends on how many pins are available Mar 25 03:55:03 https://plus.google.com/108019867520491227967/posts/LtgjjHn4ADM Mar 25 03:55:44 16 theoretically Mar 25 03:57:18 like for blue?? Mar 25 04:01:23 no idea Mar 25 04:01:53 no schematic or reference material available Mar 25 04:06:46 m_w: and one more thing.. Mar 25 04:06:49 regarding base dts.. Mar 25 04:08:22 will there be different dts for different peripherals on beaglebone blue? Mar 25 04:08:33 m_w: or will there be a base dts? Mar 25 04:12:09 base probably Mar 25 04:12:21 but what about the wireless dts?? Mar 25 04:12:25 like mentioned for the green? Mar 25 04:12:38 is wireless standard on the board? Mar 25 04:12:48 yeah.. Mar 25 04:12:59 that's what is mentioned on the blue page.. Mar 25 04:13:02 even ble is there.. Mar 25 04:13:37 where is said blue page? Mar 25 04:14:05 https://github.com/beagleboard/linux/blob/4.1/arch/arm/boot/dts/am335x-bonegreen-wireless.dts Mar 25 04:14:18 https://beagleboard.org/blue Mar 25 04:15:07 or is it via USB?? Mar 25 04:15:37 is the wifi module onboard? Mar 25 04:17:17 if it is USB then the overlay makes more sense Mar 25 04:18:32 jkridner: ping! Mar 25 04:19:49 m_w: I think it is onboard. Mar 25 04:20:02 then the overlay makes more sense right?? Mar 25 04:20:12 if it is usb.. why do you need an overlay for that? Mar 25 04:21:23 actually that doesnt make sense Mar 25 04:22:23 it should just need a wifi kernel module Mar 25 04:22:33 ok.. Mar 25 04:24:58 I think the green wireless is a population option or something Mar 25 04:25:10 and hence the different top level dts Mar 25 04:26:41 you know what a population option is? Mar 25 04:28:51 the parts are selectively populated on the PCB for different board options Mar 25 04:29:10 for cheaper boards the wireless is left off for instance Mar 25 04:30:29 so many toplevel dts files in the beagle kernel tree Mar 25 04:33:43 m_w: OK.. I get that.. Mar 25 04:33:51 m_w: can you tell what this is for?? Mar 25 04:34:01 https://github.com/beagleboard/bb.org-overlays Mar 25 04:34:29 I can see that it has got something to do with the overlays.. but what?? does it generate overlays? Mar 25 04:40:52 not sure what the deal is there Mar 25 04:41:41 I think it builds overlay firmware binaries Mar 25 05:14:24 hmmm Mar 25 05:35:13 ds2: can you please go through my proposal? Mar 25 05:45:34 hey everybody, kind of late to the party here I know... I'm interested in the sonic anemometer project. I've been talking to nerdboy over the last several days about it, and have a draft for the core of the proposal that I'd really appreciate feedback on. https://docs.google.com/document/d/1MEx5eXVJNEdANsgICRpYv12qtIIKDZ-8NK3cDtr5EL4/edit Mar 25 06:04:36 a lot of anonymous animals looking at that draft... Mar 25 06:05:05 looks pretty good so far, address any comments, add some detail Mar 25 06:05:09 and some tests Mar 25 06:05:14 you can probably chalk a couple of them up to me... Mar 25 06:06:43 alright great, thanks for taking a look Mar 25 06:08:53 can you throw in some napkin diagrams? Mar 25 06:08:57 certainly Mar 25 06:10:21 actually these beagle-y devs might ask for more detail on the pru piece Mar 25 06:10:57 are you just using the 5 ms clock timing? Mar 25 06:11:36 Is there additional jobs are required? Mar 25 06:11:45 After ”Proof of Enrollment“ and “Your proposal” Mar 25 06:12:42 5ns, yeah mostly. I was thinking about trying to do some of the simplest/earliest signal processing on them too (probably just threshold, or something like that) Mar 25 06:12:54 er, 5 nsec timing... Mar 25 06:14:23 5 nsec is admittedly lower than required, but timing does need to be below 100 nsec if we want to have a chance at .1 m/s wind speeds Mar 25 06:16:09 need to do the math and poke at some algorithms before things settle i think Mar 25 06:16:39 if you have some initial estimates/reqs put 'em in the draft Mar 25 06:16:49 like that one Mar 25 06:17:00 they're mostly in there, just far more implicit than they should be Mar 25 06:17:11 I'll make sure they're obvious Mar 25 06:17:16 minimum threshold and performance target Mar 25 06:18:04 the largest problem I"m having with that is that I can't find any information on the error in the temp/humidity corrected speed of sound equation Mar 25 06:19:31 need to do uncertainty/error analysis Mar 25 06:19:56 I mean, I can figure out how temp and humidity errors impact the result of that calc, but that calc itself isn't exact Mar 25 06:20:32 for now significant digits in == significant digits out Mar 25 06:21:06 do the calculations in full precision then figure it out Mar 25 06:21:59 I mean, the temp/humidity corrected calc is still an estimation. it makes some assumptions about the gas composition, for example Mar 25 06:22:30 gamma is specific heat ratio Mar 25 06:22:45 Visaoni: are you a ME? or know ME folks? Mar 25 06:23:19 you know the molecular weight, you could do a combining rule for MW of the mixture based on humidity Mar 25 06:24:02 then estimate gamma from MWmix Mar 25 06:24:04 ds2: I am not a ME. can't say I know any offhand, but I should be able to find some Mar 25 06:25:12 Visaoni: hmmmm teh folks that would know better on modeling is a fluid dynamics person (ME area) Mar 25 06:25:31 modeling what exactly? Mar 25 06:25:47 * nerdboy puts on applied math hat Mar 25 06:26:03 modeling how changes in humiditys/speed effect propagation speed of sound Mar 25 06:26:11 air is a fluid for this purpose Mar 25 06:26:27 there are some complicated equations that govern it Mar 25 06:26:44 I suspect to get an estimate, you need absolute humidity not relative humidity Mar 25 06:27:00 which inturn may change the density of air Mar 25 06:27:51 you can get specific humidity Mar 25 06:28:06 see - https://en.wikipedia.org/wiki/Speed_of_sound Mar 25 06:28:06 which is what you need Mar 25 06:28:06 [WIKIPEDIA] Speed of sound | "The speed of sound is the distance traveled per unit time by a sound wave propagating through an elastic medium. In dry air at 20 °C (68 °F), the speed of sound is 343.2 metres per second (1,126 ft/s; 1,236 km/h; 768 mph; 667 kn), or a kilometre in 2.914 s or a mile in 4.689 s.The speed of sound in..." Mar 25 06:28:14 that touchest on most of what I was guessing on Mar 25 06:28:24 there is a formula there for that Mar 25 06:28:32 ignore me if that has been read and discarded :D Mar 25 06:28:39 that's what the mixture molecular weight is for Mar 25 06:28:51 mixture of dry air and water vapor Mar 25 06:29:47 specific humidity => combining rule => MW of moist air => new value of gamma for moist air Mar 25 06:30:14 seems like a correct factor right there Mar 25 06:33:29 default gamma=1.4 is for dry air Mar 25 06:34:34 in they olden days they precomputed a bunch of that stuff and used lookup tables Mar 25 06:41:06 Hey Wormo Mar 25 06:41:11 you there ? Mar 25 06:41:46 ah... Mar 25 06:55:38 she had sblug mtg tonight... Mar 25 06:57:53 oh .. ohk . Mar 25 06:58:01 nerdboy, You free now ? Mar 25 06:58:20 sort of... Mar 25 06:58:39 please have a look at the proposal .. : https://docs.google.com/document/d/1GfnoWusWUeH2VsLbt1qJaRi20uNxCC8O_Xw1CHoGY9A/edit?usp=sharing Mar 25 07:00:37 nerdboy :Also, in this project, we do not require very high data transfer rates between PRUSS and ARM. So, possibly, we can use the ARM to directly read the PRUSS memory areas. Mar 25 07:00:46 (through DMA maybe) Mar 25 07:02:11 but.. my first approach will be to get everything working by using a single big module ... right from PRU to IIO subsystem. and using the proven approach of using PRUn to sample and PRUn+1 to transfer data. Mar 25 07:02:39 and then after testing other approaches . Does that make sense ? Mar 25 07:02:56 Will that be the right approach ? Mar 25 07:03:40 sounds good Mar 25 07:03:43 And later .. to divide it in modules. Mar 25 07:04:00 hey ds2 ! did you read the doc ? Mar 25 07:04:02 jsut make sure you tag GSoC time frame and post GSoC Mar 25 07:04:11 ZeekHuge: sort of :/ Mar 25 07:04:30 saw the comment messages (they get emailed out as you mark them off) Mar 25 07:05:29 " tag GSoC time frame" - you mean divide the work into weeks and evaluation periods .. right ? Mar 25 07:05:58 anything you can't fit goes to post-GSoC "sprint" Mar 25 07:06:11 oh ! ohk post GSOC and during GSOC tasks Mar 25 07:06:15 got that. Mar 25 07:08:51 :) Mar 25 07:09:59 ds2, the mails don't have the corrected/changed content .. right ? Mar 25 07:13:49 looks better, keep working on schedule/milestones Mar 25 07:13:59 don't forget testing... Mar 25 07:14:11 yeap ohk :) Mar 25 07:15:29 if you want to rate it ? out of 10 ? ( you can definitely ignore this question if you want to ) ;) Mar 25 07:21:18 ZeekHuge: just the comment box stuff Mar 25 07:36:09 Morning jic23 ! Mar 25 08:02:42 Not really here... Mar 25 08:02:55 Goofing morning though! Mar 25 08:19:01 Gag auto correct on phone. Good not goofing! Mar 25 08:19:33 Oh ffs gah not gag... Mar 25 11:13:00 is there a list of the Blue components anywhere? ie. which barometer is included? Mar 25 11:28:39 I want to work in Process Sensor Data in Real-Time. How do I go about writing a proposal? Mar 25 11:28:52 * Abhishek_ checks in Mar 25 11:29:24 Guest66286, wait till 00:30 , 26th March, maybe ? Mar 25 11:29:33 Guest66286, just kidding :P Mar 25 11:30:12 Hey Abhishek_ Just wanted to clarify one thing, SPI master driver communicates with the SPI core,which in turn communicates with SPI subsystem right? Mar 25 11:30:29 Abhishek_, ... did you see my proposal ? Mar 25 11:38:31 m_w, Just wanted to ask,SPI subsystem(via ioctls) talks to SPI core which calls on the spi_transfer function.This calls the SPI master which in turn talks to RPMsg Mar 25 11:38:48 Am I right in my reasoning? Mar 25 11:40:04 Hey bradfa, you there ? Mar 25 11:43:44 Abhishek_: can I use https://github.com/pantoniou/testpru/blob/master/testpru0.c as a userspace driver communicating with the kernel driver for pru-servo? Mar 25 11:44:23 kiran4399: That driver is only valid for the 3.8.13 kernel versions, hasn't been ported to 4.1 yet. Mar 25 11:44:54 Abhishek_: so what do you think I can do for the 4.1 kernel.. Mar 25 11:44:59 you'll have to port it to 4.1 Mar 25 11:45:04 how? Mar 25 11:45:15 Well, it's your proposal Mar 25 11:45:35 You have to figure it out. Mar 25 11:46:25 What about shubhangi's work last year?? Mar 25 11:46:28 can I use libpru? Mar 25 11:46:39 Abhishek_: ^^ Mar 25 11:46:57 that was on the 3.14 kernel, you really have to work with the latest version of the kernels. Mar 25 11:47:17 Previous work can be borrowed and built upon, but you really have to get it working Mar 25 11:48:24 so make sure you make an allowance for it in your proposal or convince us you can get the job done in the timeframe you provide us in your proposal Mar 25 11:49:59 Abhishek_, is it a good idea to use an external oscillator, to clock the 16bit parallel capture ( on PRUs ) and to clock the ADC ? or should configure PRU to provide the clock ? Mar 25 11:50:32 ZeekHuge: both ways are known to work Mar 25 11:50:40 ZeekHuge: hi Mar 25 11:50:51 hey bradfa .. Mar 25 11:51:06 so you were suggesting to use an external clock ? Mar 25 11:51:14 ZeekHuge: overall, your googledoc seems pretty good, make sure you put in a proposed schedule Mar 25 11:51:27 Thank you :) Mar 25 11:51:37 Yep working on it. Mar 25 11:51:51 ZeekHuge: it seems like you have a lot of learning to do prior to the start of gsoc, if you can more clearly define the steps there so you can know when week1 starts if you've accomplished everything, that would be good Mar 25 11:52:23 ZeekHuge: like have explicit tests your will run, then show they fail, then fix the code, such that you show you understand how the PRUs work in, for example, beagle logic (as you use that example a lot) Mar 25 11:52:32 * bradfa has to get coffee, brb Mar 25 11:55:05 Yes, timeline still needs a lot of working on it. Mar 25 11:55:25 there's very little time left now, 7 hours. Mar 25 11:55:53 yeahh... ahh Mar 25 11:56:28 so bradfa .. the thing you said there about the clock . Would you suggest using an external clock ? Mar 25 11:56:38 like an oscillator ? Mar 25 11:57:08 as a source for both, parallel capture and the ADC ? Mar 25 11:57:19 kiran4399: You've defined a new task for every week, which can get very tight. Mar 25 11:58:13 what if your timeline screws up if you find something is not working as expected? Mar 25 11:58:39 Abhishek_: Actually the thing.. most of them are already implemented.. I just have to reuse the things.. Mar 25 11:58:56 for example for the Beaglebone APIs, I borrowed the entire thing from strawson APIs.. Mar 25 11:59:18 I just have make these apis communicate with the kernel drivers. Mar 25 12:00:13 as for the ros support.. I just need to use the same apis created in week4 and 5 and write a node source code which is about 200 lines Mar 25 12:02:20 I'd just recommend that you keep a reserve week for any unseen problems you might face, then. Mar 25 12:02:31 ZeekHuge: I'd suggest using an ADC that outputs a clock along with the data so that the receiver knows when the data is valid to be sampled Mar 25 12:02:50 ZeekHuge: or yes, your statement of having an external clock clock both the PRU and the ADC Mar 25 12:03:15 Ohk .. :) back on the doc. Mar 25 12:07:58 m_w: iirc the blue has wifi over SDIO (you were talking about it last night) I think Mar 25 12:08:10 its a wl18xx module I think Mar 25 12:08:14 bradfa: A WiLink module isn't it? Mar 25 12:08:46 Abhishek_: I think so, yes, but I have no docs (but really really want to get my hands on that new am33xx module!) :) Mar 25 12:08:58 yeah, me too :) Mar 25 12:09:34 depending on the pricing and ballout of the module, I may have a lot of ideas on using it :) Mar 25 12:10:25 Abhishek_: so which one essentially needs to be ported?? remoteproc or prussdrv?? Mar 25 12:10:55 kiran4399: prussdrv can be gotten working, but it will be a non-standard kernel interface then Mar 25 12:10:59 kiran4399: isn't remoteproc "done" and in mainline? Mar 25 12:11:19 bradfa: remoteproc is but pruss_remoteproc... is another story Mar 25 12:11:38 Abhishek_: on the Linux side, what's special about the pruss_remoteproc? Mar 25 12:11:44 * bradfa is not super familiar with pru yet Mar 25 12:11:57 bradfa: Nothing more than it uses a standard kernel interface Mar 25 12:12:41 Abhishek_: wouldn't porting TI's SDK kernel (or older bb.org kernel) to use any mainline changes be all that's required there? Mar 25 12:12:43 bradfa: The UIO driver is too simple, it just provides the PRU registers to userspace and then libprussdrv does all the heavy-lifiting Mar 25 12:12:54 from userspace Mar 25 12:13:20 Abhishek_: oh ok, isn't remoteproc just a way to load firmware, start, and stop an embedded processor? Mar 25 12:13:25 bradfa: yep, but the interface of pruss_remoteproc has been changing time to time Mar 25 12:13:41 bradfa: yep, there's also the virtio and vring functionality that it provides Mar 25 12:13:55 Abhishek_: ah! ok, the vring is the part my brain is missing Mar 25 12:15:05 interface of pruss_remoteproc ? does it declare any APIs ? I wasnt able to find it in the code ? I thought it just creates those sysfs entries. And libpru then uses them .. Mar 25 12:15:13 isnt it ? please correct me .. Mar 25 12:15:33 Abhishek_, ^^ Mar 25 12:17:54 ZeekHuge: go through the kernel driver code again. Mar 25 12:20:25 Abhishek_ : Will do ... but just one question.. Are their APIs it exports ? I spent 5-6 trying to understand the code but, but I wasnt able to find any in-kernel API exporting macro. Mar 25 12:35:48 ZeekHuge: as you'd be building a subsystem, so it's up to you to figure out which symbols should get exported if so needed to. Mar 25 12:36:16 You have to/ Mar 25 12:36:42 yes ofc ! That is why I was investigating into pruss_remoteproc. And there was not a single API. Mar 25 12:37:12 It just gave some sysfs entries there .. Mar 25 12:37:28 nothing else .. and that is what libpru uses. Mar 25 12:38:01 That is why it is not generic. And no kernel module can actually use it from inside. Mar 25 12:40:37 So it is just a straight, "from PRUs to userspace" kind of, single big driver. Not in form of reusable chunks. Mar 25 12:41:12 reusable = in kernel space. Mar 25 12:41:33 ZeekHuge: so you think you can do that as a part of your proposal? Mar 25 12:42:29 Not exactly .. But pruss_remoteproc is the best starting point I can have. Mar 25 12:43:21 Basically, in first stage I will use pruss_remoteproc and develop a single driver to do the parallel capturing Mar 25 12:43:26 why hot? Mar 25 12:43:29 *not? Mar 25 12:44:03 So that we always have something that captures data. Mar 25 12:44:22 and it will also be helpful in debugging Mar 25 12:44:30 We will then break it into parts. Mar 25 12:44:38 the iio_driver Mar 25 12:44:46 the parallel_offload subsystem Mar 25 12:45:00 and the pru_parallel_interface Mar 25 12:45:13 If you want to mainline it, you need to have a well-defined interface made up of exported symbols. Where's that? Mar 25 12:45:14 so yes there will be inkern APIs Mar 25 12:45:57 Well, Its somewhat explained in the diagrams ? and the text too ? Mar 25 12:46:01 there on the doc ? Mar 25 12:47:11 And there will be further discussion on that ... Though the aim of the project is not to mainline things Mar 25 12:47:43 its to make them more usable, and doing them the linux way.. Mar 25 13:01:23 Careful not to undershoot, zeek. Mainline may not be within timeline but is necessary to be of more than minor use. RCN kernel is perhaps a fair target, but that rproc end will require agreement on interface api's for that. Other projects in and out of gsoc will overlap in that area. Something to perhaps run in parallel with whole of the rest of the project. Mar 25 13:07:22 Doing it the Linux way means everything is ready for mainline anyway :) Mar 25 13:08:07 Otherwise you are doing it then vendor bsp way! Mar 25 13:09:45 Be good to get some experience on the full kernel review process as well even if you have dependencies that need to be upstreamed first. Mar 25 13:10:23 Anyhow end of train journey for me so bye for now. Mar 25 13:11:29 +1 to what jic23 said. Mar 25 13:12:03 You don't want to "just get things working" Mar 25 13:12:14 (because they already are :) ) Mar 25 13:16:13 Last comment.. Proposal should highlight need to spend time on this not try and specify to tightly what they are. May be talking 10 + review cycles to get that right! Mar 25 13:17:57 yep, considering it's only less than 6 hours left, but you must get a timeline in Mar 25 13:21:02 Hey Abhishek_ ! Can u comment on my project proposal ? Mar 25 13:21:22 I kind of missed the " comments from the mentors " question! Mar 25 13:21:47 My proposal : https://docs.google.com/document/d/1WDq5zcQAIUx6BA3M5C4u7m_gzjdtIAhn28mI-8iMgkQ/edit Mar 25 13:22:55 ds2 : online ? Mar 25 13:23:00 nerdboy : online ? Mar 25 13:23:14 yashoza: They'll come online probably by 9-10pm Mar 25 13:23:21 9-10pm IST that is Mar 25 13:23:38 they're in the Pacific Time zone Mar 25 13:23:49 yeahh..would be a little late :( Mar 25 13:24:03 Is it possible for u to comment ? Mar 25 13:24:22 have to be afk for a while, lemme see. Mar 25 13:24:44 sure! Mar 25 13:25:21 :) Mar 25 13:42:08 hai Mar 25 13:42:32 hi Mar 25 13:43:47 i want to know how do you select an abstract Mar 25 13:45:32 rather than using beagle board is there any specific iot to be used Mar 25 13:47:29 can any one answer please Mar 25 13:53:06 hai Mar 25 14:06:53 Hello peopl Mar 25 14:06:54 ee Mar 25 14:07:04 anyone there on PRUs? Mar 25 14:11:51 what do you want to know? Mar 25 14:22:25 Sorry to disturb i was looking for some feedback on my proposal Mar 25 14:22:57 https://groups.google.com/forum/#!topic/beagleboard-gsoc/HRiaKBzyg4Q Mar 25 14:43:38 hello Mar 25 14:43:43 anyone Mar 25 14:54:35 Abhishek_ : are you there? Mar 25 14:54:56 yup Mar 25 14:54:59 im still here Mar 25 15:08:01 sarat: you aren't a mentor -_- Mar 25 15:08:31 yup i am seeking mentor Mar 25 15:08:35 :( Mar 25 15:08:46 i thought you are a mentor my bad. Mar 25 15:09:41 * jkridner peeks in for last minute requests. Mar 25 15:09:50 which hour is the deadline? Mar 25 15:10:19 4 Hours to go Mar 25 15:10:20 3 hours .. 49 minutes and .. 40 second Mar 25 15:10:23 from now Mar 25 15:10:42 that one is correct Mar 25 15:12:36 :-) Mar 25 15:12:57 are all mentors signed up on withgoogle.com? Mar 25 15:16:31 alexhiam: I updated the blinky firmware on GitHub...complete with makefile and stuff :D Mar 25 15:17:24 Please do go through my code for host linux.....I am unsure if I need to initialize interrupts or not....coz I'm not using any (I have to initialize events even though I don't need em) Mar 25 15:17:39 udayansinha: what is your github link? Mar 25 15:17:50 * jkridner is just peeking in on stuff. Mar 25 15:17:56 https://github.com/UdayanSinha/GSoC_Beagleboard Mar 25 15:18:36 udayansinha: why the build and clean shell scripts? Why not do that with the Makefile? Mar 25 15:18:45 * alexhiam wonders if that CCS weirdness Mar 25 15:19:22 Nope....I referred a tutorial for the scripts and files....I maintained the structure Mar 25 15:20:11 I have to make another change too...they're not running the code at address 0....they have changed it and used a custom routine to write code at another location......I gotta fix that too (need a board to test all this stuff) Mar 25 15:21:44 I already cleaned up quite a bit of the host code...removed stuff i didn't need Mar 25 15:23:00 k, a reference to that tutorial in there would be good Mar 25 15:23:12 cool will add it to the readme Mar 25 15:23:39 this is the one: https://www.embeddedrelated.com/showarticle/603.php Mar 25 15:23:57 have you looked at the TI workshop stuff? Mar 25 15:24:17 http://processors.wiki.ti.com/index.php/PRU_Training:_Hands-on_Labs Mar 25 15:24:23 yeah I did....they're using a JTAG emulator in between.... Mar 25 15:24:45 that's why they don't need a linux loader program....that's what confused me yesterday Mar 25 15:24:53 gotcha Mar 25 15:25:27 I just used the TI stuff to develop blinky Mar 25 15:27:21 linux loader from the other tutorial Mar 25 15:39:22 if anyone happens to be here who submitted a proposal as just a final pdf, we can't see it until the deadline so we can't give you feedback. If you want a chance to get suggestions and edit accordingly, you need to upload a draft google doc ASAP! Mar 25 15:41:08 and if your google doc doesn't give edit permissions then we can't leave comments! Mar 25 15:41:08 Should one have to include the cross compilation code( github link) in the final pdf? Mar 25 15:42:09 that would be good Mar 25 15:44:15 Thanks Mar 25 15:51:57 to which mail id do i have to share my abstract Mar 25 15:54:08 gauti: what do you mean? It should be on summerofcode.withgoogle.com Mar 25 16:10:13 * alexhiam goes afk for a bit... Mar 25 16:22:17 jkridner: Did you check your email? Mar 25 16:27:26 jic23, ? there ? Mar 25 16:38:13 ds2 : available ? Mar 25 16:38:20 nerdboy : available ? Mar 25 16:40:04 yashoza? Mar 25 16:42:22 got a mtg coming up... Mar 25 16:43:22 Hey! Can u tell me your views on my project, and how it would help the community? I gotta include views from my mentors in the proposal Mar 25 16:43:24 ! Mar 25 16:45:18 it benefits the community because you're creating something useful that doesn't exist Mar 25 16:47:53 Okkay Mar 25 16:48:04 at least some group of beagle-y users that need it... Mar 25 16:49:38 Yup Mar 25 16:52:35 Hey m_w . If successfully completed, what will its impact be on the BeagleBoard.org community? Mar 25 16:52:41 its = my project Mar 25 16:53:02 need to include it in the doc Mar 25 16:53:08 hey jic23 ! you there ? Mar 25 16:53:40 Yes now. Mar 25 16:53:57 need to include ans to this : If successfully completed, what will its impact be on the BeagleBoard.org community? Mar 25 16:54:12 quoting from community members. Mar 25 16:54:30 jic23, can you please give your reviews on that ? Mar 25 16:54:38 i have completed my proposal anyone can verify it Mar 25 16:55:02 can u please give your review Mar 25 16:55:40 Abhishek and others commented on usefulness in particular for showing people how to do stuff 'properly' Mar 25 16:56:10 Hmm, from me is slightly different angle. Mar 25 16:56:19 Abhishek_, ^^ Mar 25 16:57:16 i need help regarding my proposal Mar 25 16:58:41 Having a single driver covering a wide range of parallel ADCs connected to a PRU equipped to processor has worth in of itself. Mar 25 16:58:42 I think the project will make PRU interaction more approachable. Mar 25 16:59:13 ohk thank you jic23 , m_w Mar 25 16:59:29 can you please give your mail id I will send my proposal to verify . Mar 25 16:59:31 ds2, ? ^^ can you please give your opinion on this ? Mar 25 16:59:52 Any progress beyond that state in breaking that driver up will provide more value still. Mar 25 16:59:54 also it shows how the BBB can perform task that would otherwise require expensive hardware. Mar 25 17:00:30 tasks Mar 25 17:01:13 this is your mail id -yashoza? Mar 25 17:02:02 Development of a full parallel offload driver and abstraction in particular may lead to generic reusable drivers for parallel ADCs on numerous systems. Mar 25 17:02:41 ajitkr: you shouldn't be emailing your proposal to anyone, it needs to be submitted to summerofcode.withgoogle.com Mar 25 17:03:09 but can u please verify it further Mar 25 17:03:09 Also, on a more practical personal point it will give a cheap entry point for testing my subsystem at the higher speed end. Mar 25 17:03:15 ajitkr: Post a link here if you want someone to check your proposal Mar 25 17:03:48 https://docs.google.com/document/d/1s58zNR9J_84xJuh1U29DUtDyjnt8aJhTwBmDv3Qai3A/edit?usp=sharing Mar 25 17:03:59 this is my link -ajit Mar 25 17:04:13 Then there are the more direct applications that might lead from this support and example code / documentation. Mar 25 17:04:25 ZeekHuge: library of examples on proper kernel API usage with the PRUSS Mar 25 17:04:43 Zeek, pick what your like to quote! Mar 25 17:05:00 ajitkr: wait, so that's different than the doc currently submitted on summerofcode.withgoogle.com? Mar 25 17:05:31 ok nvm, I see it's the same Mar 25 17:06:01 However I only count as community in a more general way. Can't really claim to be an active member of the bb community :) Mar 25 17:06:11 Everything ;) Mar 25 17:06:13 ZeekHuge: will lead to an API framework that allows other drivers to "plug in" to remoteproc and do their thing Mar 25 17:06:41 Thank you Abhishek_ , ds2 :) Mar 25 17:06:58 yes i modified it Mar 25 17:07:08 +1 on that part though that may be true for the 3 bus project as well. Mar 25 17:07:48 ajitkr: it's still not clear to me how it would tie in with last years project. It still seems like it will be redoing the work already done Mar 25 17:09:07 the half part of the project same as the last year but other half is different Mar 25 17:09:24 alexhiam, would you like to say something on that ? Mar 25 17:09:33 ajitkr: but doesn't that mean half the project is already done? Mar 25 17:10:23 so please give me what is your view about my project Mar 25 17:10:36 please tell me Mar 25 17:10:55 STUDENTS: Don't forget to submit a final app to the GSoC site Mar 25 17:11:21 ajitkr: well, that's the big thing. It seems like you're proposing to redo from scratch everything that was done in the previous project. If that's not the case it's not clear in the proposal currently Mar 25 17:11:59 it's about the same as last night, no significant changes really Mar 25 17:13:08 what should be change in my proposal tell me further Mar 25 17:13:39 we can't tell you what to write... Mar 25 17:14:12 but u can tell me what should be change. Mar 25 17:14:59 sounds like too much overlap with last year's project, how much did you study that one? Mar 25 17:16:01 ajitkr: well redoing a project from last year doesn't work, so you'd want to make sure you understand what that project currently does and make it clear in the proposal that you would be building off of that Mar 25 17:16:17