**** BEGIN LOGGING AT Thu Mar 19 02:59:58 2015 Mar 19 04:07:29 nerdboy : ping Mar 19 04:10:10 sidbh_:ping Mar 19 04:14:48 i'm facing this strange issue when i'm using the PRUSS-C library Mar 19 04:15:10 I was trying to run ePWM module using PRUSS Mar 19 04:15:51 but couldn't get it to run Mar 19 04:16:27 so i tried to run a modified version of the blinkled test Mar 19 04:17:09 this is the file : https://github.com/altairpearl/starterware_pru.git Mar 19 04:17:26 wrong one : this is the file https://gist.github.com/altairpearl/c0b7fae3a86ec6870572 Mar 19 04:18:42 the program doesn't run as expected . Ideally it should blink each pair twice and stay on but what it does is that it blinks one pair and keeps on looping that Mar 19 04:19:28 this happens when i try to define the function outside the main and try to invoke it within the main. if I define everthing within the main it works fine Mar 19 04:19:31 any help! Mar 19 05:14:49 sidbh_: ping Mar 19 05:25:10 nerdbpy:ping Mar 19 05:25:21 nerdboy:ping Mar 19 06:57:57 altairpearl, pong Mar 19 07:01:39 sidbh_: hey Mar 19 07:01:40 i'm facing this issue Mar 19 07:01:49 my code :https://gist.github.com/altairpearl/c0b7fae3a86ec6870572 Mar 19 07:02:10 this was a test . i was actually trying to run epwm using pruss Mar 19 07:02:40 this could should ideally make a pair of userleds to blink twice and stay on Mar 19 07:02:47 code* Mar 19 07:03:09 but what it does is it just keeps on blinking one set of leds alone Mar 19 07:03:22 when i include all the loop within the main code it works fine Mar 19 07:03:24 altairpearl, stop the USR LEDs blink inside userspace Mar 19 07:03:36 btw Mar 19 07:03:41 i have succesfully done that Mar 19 07:04:13 i have disabled them in the device tree itself Mar 19 07:05:14 this code is working as expected : https://github.com/altairpearl/PRU-C/blob/master/ARMLinuxA8/blinkledv0/blinkled.c Mar 19 07:05:26 ok, then initialise them, set GPIO DIRs etc, that might help Mar 19 07:06:16 i have done that Mar 19 07:06:52 what i'm facing is when i use a separate function to blink the leds and invoke them within the main this happens Mar 19 07:07:12 but when i do that within the main it works perfectly Mar 19 07:07:44 ohhh... Mar 19 07:08:24 yea :( Mar 19 07:08:47 where are you compiling ur code Mar 19 07:08:53 on BBB Mar 19 07:10:22 kiran4399 got some issues too when he compiled onboard but the same code worked when compiled offboard, try compiling on x86 machine and see what happens, it might be a build issue Mar 19 07:11:01 oh okay . Ill try that Mar 19 08:55:43 morning! Mar 19 09:59:56 nerdboy : ping Mar 19 10:28:19 'Morning Mar 19 10:28:40 mornin FREEZX Mar 19 10:29:50 I wrote about my idea last night, would appreciate getting feedback on it Mar 19 10:30:12 will take a look later today Mar 19 10:30:20 Thanks Mar 19 12:34:58 Good day. I am a Computer Science Undergrad. Student in Germany, and I am interested in ADC improvement in Beagle Boards. Mar 19 12:36:40 hi Mar 19 12:37:01 Hello. Mar 19 12:37:25 I want to know what tasks I can fulfill , before submitting my proposal to the GSoC form. Mar 19 12:38:03 there is a qualification task Mar 19 12:38:06 did you do that? Mar 19 12:38:07 Also, advice and links to already-existing code and schematics could be of great help. Mar 19 12:38:42 you arew referring to http://elinux.org/BeagleBoard/GSoC/Ideas#Enhance_ADC_driver_for_BeagleBone_and_BeagleBone_Black ? Mar 19 12:38:53 the proposed mentor is ds2, but he is asleep atm Mar 19 12:39:03 can you check back later in the day as he is on US west coast Mar 19 12:39:31 Yes, I am referring to that. Thank you for information regarding his location, I will come back later. As for the task, I'll complete it today. Mar 19 12:39:52 he is also on the bbb-gsoc mailing list Mar 19 12:40:02 so you might write him a short email Mar 19 12:40:08 and willkommen Mar 19 12:40:28 abhishek_:ping Mar 19 12:40:46 Good, thank you! Have a good day. Mar 19 12:41:11 karki alexanderhiam Abhishek_ :Submitted my proposal for PRU-Bridge yesterday and would love to hear your comments :) Mar 19 12:41:26 ok Mar 19 12:41:53 I also submitted my own idea and i need feedback Mar 19 12:43:04 I volunteered as a mentor for the project, I'm hoping Abhishek_ and alexanderhiam will also review and volunteer :) [need to discuss a few specifics with them] Mar 19 12:44:05 oh ok cool :) Mar 19 12:45:43 karki : hey can you tell me more about the project "PRUSS support for newwer kernels" ? Mar 19 12:46:37 there is already a rproc framwork already being used in beaglelogic Mar 19 12:49:30 Abhishek_ : ping Mar 19 12:51:49 Yes there is Mar 19 12:54:23 so in this project are we going to write a similar framework or rewrite it so it can patched to linux kernel ? or am I getting this wrong ? Mar 19 12:59:56 i have some knowledge about the PRU architecture and i am familiar with programming it using assembly Mar 19 13:09:01 jkridner: Are you aware of the fact that, for pruduino, i am not only configuring c9 with it.. but also building live web-pages for it ?? (just like bonescript) .. Mar 19 13:09:25 jkridner: It will be more user-friendly... Mar 19 13:09:43 "configuring c9 with it"? do you mean providing a 'runner'? Mar 19 13:10:14 and, wouldn't it just be a JavaScript library anyway, so that you just run a JavaScript program? Mar 19 13:10:25 I think you are overstating "configuring c9 with it". Mar 19 13:10:48 do you mean "in addition to being able to run programs in c9..."? Mar 19 13:11:07 providing the rpc mechanism like what bonescript does is good. Mar 19 13:11:36 or just using bonescript's mechanism. Mar 19 13:11:57 anyway, I like what that you are thinking about how the end user experience will be to learn it.! Mar 19 13:14:07 jkridner: I have submitted a proposal for the PRU-Bridge project on melange.Would love to hear your views and comments on it :) Mar 19 13:15:34 Can i get feedback on my project idea https://groups.google.com/forum/#!topic/beagleboard-gsoc/CC2CZ6pmuV4 Mar 19 13:15:41 I'll be checking in on melange and the mailing list this afternoon. Mar 19 13:15:42 or proposal on melange Mar 19 13:15:46 ok thanks Mar 19 13:15:55 Thanks :) Mar 19 13:17:10 yes.. yes you are absolutely right !!! what I meant when I said configuring c9 with it is that the end-user can just run the code as a js program in c9 ide... in addition to that.. he can also use the live-web pages Mar 19 13:17:20 jkridner: ^^^ Mar 19 13:17:56 jkridner: remote procedure calls will be an awesome method for this.. Mar 19 13:20:05 jkridner: I am sort of combining my ideas with noduino and bonescript and comming out with these new ideas.. Mar 19 13:20:37 Abhishek_:ping Mar 19 13:21:17 jkridner: I also want to build GUI for input devices... on the web-pages.. Mar 19 13:23:47 jkridner: like.. you know... virtual pot's, buttons etc.. where he can give input to beaglebone VIRTUALLY from the computer.. what do you think about this idea ?? Mar 19 13:29:05 Ideas are cool. Execution matters more than ideas Mar 19 13:31:25 Abhishek_: thanks.. Mar 19 13:32:33 Abhishek_ : am I asking the wrong questions ? just point me to where I should look more please Mar 19 13:33:44 Look at the BeagleLogic and PRUSpeak drivers and previous discussion threads on the list Mar 19 13:34:37 hmm ok Mar 19 13:39:00 kiran4399: What I meant to say was that it is nice to be able to come up with new ideas but must also in parallel think on how to be able to deliver it within the GSoC timeframe Mar 19 13:40:03 Abhishek_: OK.. Mar 19 13:42:17 Always a trade off between BFS and DFS. Mar 19 13:43:38 Abhishek_: good one !!! :D Mar 19 13:48:39 * jkridner gets scared by the word "virtual" Mar 19 13:49:38 kiran4399: what are you thinking to do with these GUI pots/buttons? Mar 19 13:49:46 have them reflect the status of the real devices? Mar 19 13:50:03 like Fritzing, but live-updated? Mar 19 13:50:09 * jkridner has to guess. Mar 19 13:50:27 av500: that guy interested in adc i think he is from my university Mar 19 13:50:32 yes Mar 19 13:50:34 same IP :) Mar 19 13:50:39 yep Mar 19 13:50:50 vvu: find him! Mar 19 13:50:54 i made some advertisment in our cs club Mar 19 13:50:58 he is 1st year Mar 19 13:53:03 jkridner: Well, users need not have hardware all the time.... it is a GUI... so as to make the live web-pages MORE user friendly for beginners !!! Mar 19 13:55:48 jkridner: they may be used in place of real hardware .. what do you think ? Mar 19 14:03:12 kiran4399: a simulator would be really cool, but I'd want to see enough of a minimal prototype to believe you can execute it. it is complicated and I'm not sure I'd be able to mentor you all the way through it. Mar 19 14:03:49 a fully simulated circuit that, when connected to a board, would then tie the physical to the simulation, that'd be cool. Mar 19 14:05:12 +1 to what jkridner said Mar 19 14:07:13 jkridner: No problem... I will work hard and get it done.. just tell me what kind of prototype you would want ? This semester in my school I made a small button simulator using bootstrap for raspberry-pi. Mar 19 14:16:59 kiran4399: I'm not sure what the UI would look like. I need a visual example and if you have dynamic items in the UI, I'd want to see something that shows you have the skills to create and object of that complexity and animate it. Mar 19 14:18:05 something like... here's an example of a simulated circuit... see how this moves... just need to tie up the hardware to it... we'll create a dozen of these as the GSoC project. Mar 19 14:18:20 vvu great Mar 19 14:22:19 jkridner: I basically thought of doing only a button, switch and a slider and put them on the live web-pages.. shall I show you a button simulator for beaglebone ? Mar 19 14:23:51 jkridner: I mean it is like.. when user presses the button .. then led will glow on the beaglebone.. a simple gui using bootstrap. what do you think ? Mar 19 14:29:29 I'm having to guess too much as to what you mean. I guess glowing is really easy, where as moving a slider is harder. Mar 19 14:29:53 * jkridner would prefer animations on canvas like BBUI does.... because I can understand it. Mar 19 14:36:06 * vvu thinks the number of proposals on melange is kinda small Mar 19 14:38:30 karki: sent clean pull request. all good now. please have a look if you have time :) Mar 19 14:38:40 closed previous one Mar 19 14:39:50 ok Mar 19 14:40:23 kiran4399 : I guarantee you that you won't have that much time on your hand. Mar 19 14:40:33 focus at what you want to do and how you want to Mar 19 14:40:45 don't keep adding features and targets Mar 19 14:41:12 a bigger proposal does not imply a better one Mar 19 14:41:39 shubhangi : have you submitted your proposal Mar 19 14:41:40 ? Mar 19 14:42:04 tonight Mar 19 14:42:08 hmm Mar 19 14:42:23 I had asked you to finish that before the pull request :p Mar 19 14:43:00 I won't be merging it before submission deadline, so doing that first is kinda pointless Mar 19 14:43:21 okay. just didnt feel right with that. was a big blunder. learnt git now Mar 19 14:43:25 :) Mar 19 14:43:30 good good Mar 19 14:43:38 was working on proposal also alongside Mar 19 14:43:38 lemme take a look then Mar 19 14:43:43 jkridner: I think I did not express my idea properly. I don't want to make these GUI's very complicated.. I want to keep it very simple. The main reason why I am going for these is to make these web-pages more user-friendly. Mar 19 14:43:55 karki: Thanks.. I Mar 19 14:44:12 karki: Thanks.. I'll take your advice. Mar 19 14:44:35 just one correction to be done. missed out on it Mar 19 14:44:35 kiran4399: as long as it is really clear in the proposal. Mar 19 14:44:57 kiran4399 : and moreover it would be nicer to just create a bone script interface Mar 19 14:45:10 RPC is already there Mar 19 14:45:26 done Mar 19 14:45:42 karki: Yes.. Mar 19 14:46:30 Can I ask certain questions regarding the ARM executable? Mar 19 14:46:34 question* Mar 19 14:47:31 Is the date that has to be printed to be the date on which I made the executable or the date at the time of the execution? Mar 19 14:50:13 karki : anything missing ? Mar 19 14:50:25 jkridner: one more thing.. regarding the interface, can you show me the rpc mechanism here: https://github.com/jadonk/bonescript ? Mar 19 14:55:05 kiran4399 : It's opensource for a reason, use the source code. Mar 19 14:55:21 did you even try figuring it out? Mar 19 14:58:27 * karki remembers that he got zero help from panto when he asked a question without struggling for the solution first! Mar 19 15:00:35 karki: best approach :) Mar 19 15:00:52 vvu : always :) Mar 19 15:01:19 vvu : In fact it actually helped me more than just being fed the solution! Mar 19 15:04:24 +1 Mar 19 15:05:50 same here, av500 pointed out the magic hexdump tool to me :) Mar 19 15:06:35 karki: u were going to paypal right? or i am mistaken? Mar 19 15:07:00 PayPal :) Mar 19 15:07:26 for this summer? Mar 19 15:07:30 yeah Mar 19 15:07:35 also have a offer there Mar 19 15:07:39 no, not for the summer Mar 19 15:07:48 uptill the summer Mar 19 15:08:00 I'm interning there right now! Mar 19 15:08:52 vvu : I'll be joining them (probably) as a FTE in june sometime Mar 19 15:09:12 I also have another interesting offer by a local startup Mar 19 15:12:38 nice! Mar 19 15:12:52 yesterday i had an interview with Palantir Mar 19 15:13:04 for fte Mar 19 15:14:52 hello guys Mar 19 15:15:02 hello there Mar 19 15:15:50 i wanted to ask abouth the idea for gsoc Mar 19 15:15:56 i have something to discuss Mar 19 15:18:22 just ask Mar 19 15:21:11 is an automated red light traffic camera powered by beaglebone that captures image using image processing an considerable idea to be applied for gsoc? Mar 19 15:23:39 the bbb is not for image processing Mar 19 15:24:11 using OpenCv Mar 19 15:26:47 open cv can be used for this purpose Mar 19 15:30:40 vvu : How'd it go :D Mar 19 15:31:11 Palantir I've heard is tougher to crack than Google Mar 19 15:32:26 it was for the infrastructure team Mar 19 15:32:41 shubhangi : Tremendous improvement :D Mar 19 15:32:41 a lot of system design questions and how would i debug stuff Mar 19 15:32:53 hmm... Mar 19 15:32:53 coding wise i had something with trees Mar 19 15:33:08 system design meaning ? Mar 19 15:33:11 u have a familly tree and u want to find lca of 2 nodes Mar 19 15:33:21 they put me to design yelp Mar 19 15:33:25 ah Mar 19 15:33:31 and what data structures i would use and why Mar 19 15:33:34 that way Mar 19 15:33:39 karki : thanks a lot :D. i have worked out how to add decimal delays. will finish it as soon as i submit proposal Mar 19 15:34:24 karki: and for one interview i had to find live bugs in a server Mar 19 15:34:34 interesting Mar 19 15:34:36 with terminal and everything Mar 19 15:34:40 it was super cool Mar 19 15:34:46 I can see why Mar 19 15:34:48 :D Mar 19 15:34:59 the teamin london is just formed by 2 peeps Mar 19 15:35:10 they are starting the IQE team here Mar 19 15:35:57 I didn't know you were a infra person, I thought of you as an embedded systems geek :p Mar 19 15:36:27 vvu : Have you worked with cloud and scalability before? Mar 19 15:36:32 mainly linux geek Mar 19 15:36:43 not that much but read a lot of stuff on that Mar 19 15:36:46 so LXCs and stuff Mar 19 15:37:04 actually i am interested in OS stuff Mar 19 15:37:33 working on OS stuff translates to zero money in india Mar 19 15:37:37 :( Mar 19 15:37:51 :'( Mar 19 15:38:14 * karki rolls and cries in a corner Mar 19 15:38:24 go to US or somewherw Mar 19 15:38:34 I have a job offer in london Mar 19 15:38:41 fintech startup Mar 19 15:39:03 mainly ML and data analytics Mar 19 15:39:12 with some server side stuff Mar 19 15:39:15 RoR Mar 19 15:39:18 interested in that? Mar 19 15:39:31 not planning to move out of india for now Mar 19 15:40:00 ah.. Mar 19 15:40:01 But what does palantir do OS related ? Mar 19 15:40:42 not much, i applied for the infra part Mar 19 15:40:53 if i get the job i will tell you :) Mar 19 15:41:38 anyway in EU i wanna go just in UK because of the language Mar 19 15:41:43 do tell me :) Mar 19 15:41:53 makes sense! Mar 19 15:42:03 How hard was the coding though ? Mar 19 15:42:26 not super hard but took 40 mins totally Mar 19 15:42:53 Legend has it that Palantir > Google ;) Mar 19 15:42:53 he just wanted a solution he did not care that much of complexity Mar 19 15:43:03 thats odd Mar 19 15:43:07 i don't think it is true Mar 19 15:43:16 I thought these people grill for complexity Mar 19 15:43:21 i know someone who got FTE without coding test Mar 19 15:43:26 just system design Mar 19 15:43:32 Well Mar 19 15:43:38 could that person code ? Mar 19 15:43:41 maybe for the soft dev team Mar 19 15:43:50 yep Mar 19 15:44:05 then probably they just saw previous projects Mar 19 15:44:10 maybe Mar 19 15:44:16 I personally find 'coding' pointless Mar 19 15:44:25 in interviews Mar 19 15:44:34 90% innovation 10% coding Mar 19 15:44:46 they really depend on how you think on that one day Mar 19 15:44:58 i know Mar 19 15:45:06 and most interviewers ask problems that have some 'trick/hack' Mar 19 15:45:16 i screwed up a bloomberg interview bcoz of this Mar 19 15:45:22 heh Mar 19 15:45:26 karki, what did you use ML for? It seems like a very interesting language and really underrated. Mar 19 15:45:31 the binary type questions either u know either u fail Mar 19 15:45:42 ML => machine learning Mar 19 15:45:45 not the language Mar 19 15:45:45 dmitescu: ml machine learning Mar 19 15:45:51 not sml :) Mar 19 15:46:16 Oh, okay. :) Mar 19 15:46:30 vvu : I'd choose palantir over google Mar 19 15:46:42 Google is way too big for my liking Mar 19 15:47:00 true that... Mar 19 15:47:02 * karki prefers startup environments! Mar 19 15:47:11 amen to that! Mar 19 15:47:17 I'm seeing it with PayPal Mar 19 15:47:23 not exciting enough Mar 19 15:47:35 what are u doing mainly there? Mar 19 15:47:56 same here thought u were embedded sys type of person Mar 19 15:48:05 I'm working on big data management and visualization Mar 19 15:48:21 Is using something like https://kazuhikoarase.github.io/simcirjs/ feasible for a basic beagle simulator Mar 19 15:48:27 ? Mar 19 15:48:37 vvu : nope, embedded is just my hobby. I'm a core CS guy Mar 19 15:48:43 if u like the area check out Quid Mar 19 15:48:45 abhishek is a h/w guy Mar 19 15:48:56 link? Mar 19 15:49:12 just google it, i'm on my phone Mar 19 15:49:22 cool startup but in SF Mar 19 15:49:35 found it :) Mar 19 15:49:45 but I'm again, into OS stuff Mar 19 15:49:51 or maybe DBMS Mar 19 15:49:56 things like that Mar 19 15:50:29 it really depends, I want to do a lot of things..... But so little time :'( Mar 19 15:51:30 same here...hi have also my BSc thesis Mar 19 15:51:56 well we are young :) Mar 19 15:52:08 help needed to make a proposal for Demo Android App Using BBBAndroid. Mar 19 15:53:09 Oh well. You get ready to help someone and he runs away. Mar 19 15:53:20 hehe Mar 19 15:53:46 pops back Mar 19 15:54:05 iqan: OK. What exactly do you need help with for your proposal? Mar 19 15:54:16 Template? Content? Schedule? Scope? Mar 19 15:54:37 actually regarding the scope.. Mar 19 15:54:56 i am thinking about CAN bus interfacing with BBBAndroid Mar 19 15:55:10 to be used for vehicles Mar 19 15:55:49 karki: need to split catch you later Mar 19 15:56:00 vvu : bye :) Mar 19 15:56:45 iqan: OK. So you'd like to produce something related to OBD using CAN. Do you have some ideas about your proposed project beyond that? Mar 19 15:57:10 iqan: How do you think your project will help the BeagleBoard community? Mar 19 16:03:35 hendersa : I am currently working with raspberry pi board and some GPIO interfacing with that.. i found this https://github.com/TrentSeed/BMW_E46_Android_RPi_IBUS_Controller. so i thought if using BBBA we can build some app and product for vehicles Mar 19 16:05:34 iqan: While building a stand-alone vehicle diagnostic device is a nice project, BeagleBoard.org is looking for something that will benefit the BeagleBoard community by either improving the software used for the BB.org hardware or helping others use BB.org hardware to its full potential. Mar 19 16:06:19 iqan: If you were to develop a project like this, what would its benefit be to BB.org and its users beyond its use as a diagnostic device? Mar 19 16:07:19 iqan: Is it an educational example or reference implementation for others wishing to build a similar device? If so, how do you propose to make it so? Mar 19 16:11:39 hendersa : I think it would be a nice start for users think about contributing towards not just the code but some hardware solutions too. Mar 19 16:12:34 No, its not but can be an educational example or reference implementation. Mar 19 16:14:46 I need feedback to help me improve my proposal Mar 19 16:15:03 iqan: We have a number of people interested in submitting proposals developing hardware interfacing libraries/examples with Android, and I agree that making such work available to the BB.org community will help encourage more hardware-related projects by making it easier for beginners to do so. Mar 19 16:16:00 iqan: But, keep in mind that the projects that are selected for funding will be selected based upon how broad the impacts of those projects will be. Mar 19 16:16:32 iqan: We want to give students an opportunity to work on neat projects, but we also need those students to help advance BB.org's goals. Mar 19 16:17:29 iqan: If you believe that you project will have broad appeal to the BB.org community and will serve as a good reference implementation/example/tutorial for others, then you should submit a proposal and give it a try. Mar 19 16:18:15 iqan: But, if the scope of your project is too narrow, then other hardware-interfacing projects will be selected instead because they will have a broader impact on our users. Mar 19 16:18:19 hendersa : I dont know impacts. But I am learning things, so I have no that broad idea about project, that's why I need help :) Mar 19 16:18:41 not* Mar 19 16:19:39 iqan: OK, think of it this way: You are looking at a variety of ARM single-board computers and you want to make a neat project. What kind of project would make you select a BeagleBone Black or BeagleBoard-X15 over another board? Mar 19 16:21:02 iqan: It may be a complex project, so would having tutorials or reference implementations of other projects help you? What kind of software would you need to help you get started? Mar 19 16:21:40 iqan: Maybe you'd need a tool that doesn't exist yet, or something in the kernel or system needs to be fixed up. Mar 19 16:22:27 iqan: There are many aspects to consider beyond just making the project. The project also has to improve the experience with BB.org hardware for those that come after you. Mar 19 16:23:29 iqan: So, if you'd like to do a project, think about what exactly you'd like to produce and how it will help others. Are you doing just software? Are you also creating a cape interfacing board? Mar 19 16:24:38 iqan: The mentors are here to guide you and answer your specific questions, but we can't hold your hand step-by-step and give you a project that is like a homework assignment with everything laid out. Mar 19 16:25:11 iqan: The best thing for you to do is to try writing a proposal and letting us make comments on it. Mar 19 16:25:45 iqan: State exactly what you'd like to accomplish and your schedule for doing so, and we can then provide you with guidance. Mar 19 16:27:32 hendersa : I have done some searching and I think I will be using : A BBB, CANBus Cape with DCAN1 Interface supporting Socket CAN, An app for BBBA , maybe app for wearables Mar 19 16:27:56 yeah.. I am currently trying to make a proposal.. It will be done soon. Mar 19 16:29:20 iqan: OK, sounds good. The more details you can provide, the better. Try to make it a big, ambitious project that can provide benefits to others as a learning example. Mar 19 16:29:37 iqan: But, be realistic about what you are able to accomplish. Mar 19 16:31:43 jkridner: I was tied up writing a proposal during the IRC meeting yesterday, so I couldn't make it. I reviewed the logs from the meeting. Anything info I need to know? Mar 19 16:32:26 hendersa : Yeah sure.. and anyother application that you can suggest where we can use other buses like SPI,I2C,etc. So I will think about those,too. Mar 19 16:33:35 shubhangi: there's improvement but this is just the beginning Mar 19 16:34:48 iqan: I think that any practical interfacing examples with those other buses is fine. Interfacing with premade component breakout boards with those buses (sensors, EEPROMS, etc) would be convenient. Mar 19 16:35:46 iqan: Remember, it is the number of people you can help with your work that will be a major factor. You don't necessarily need to have one big project that uses every type of bus together (though you certainly could). Mar 19 16:36:01 there are deep and turbulent waters ahead :) Mar 19 16:38:07 hendersa : okk..need to read more now.. thanks for the help. Mar 19 16:39:23 Can i get some feedback on my idea? Mar 19 16:39:23 https://groups.google.com/forum/#!topic/beagleboard-gsoc/CC2CZ6pmuV4 Mar 19 16:45:16 freezx: I think it looks like a very nice idea for a project, particularly if it were integrated into bonescript. It isn't my area, though, so I will defer to others for more detailed comments and direction. Mar 19 16:45:34 hendersa: just want you to engage with students as much as you can. Mar 19 16:47:07 jkridner: Got it. Mar 19 16:48:57 hendersa: Thanks for the input Mar 19 16:50:20 freezx: I think a good application of your work would be to integrate the simulator into the existing bonescript setup so that the simulator would be immediately accessible when the BBB is first plugged in, like bonescript is. Mar 19 16:51:18 freezx: Like you can develop bonescript on one side of the screen and the circuit on the other. Get it how you like it and then "flip the switch" to go from simulated execution to real execution on the hardware. Mar 19 16:52:20 hendersa: that's exactly what i'm tried to say in my second mail Mar 19 16:52:25 *i tried Mar 19 16:53:38 So like you can create the circuit in the simulator, and once you like what you have, connect to the beaglebone and run it on the board Mar 19 16:54:04 * hendersa reviews the proposal on Melange. Mar 19 16:54:19 Would this library be sufficient to supply the basic elements? Mar 19 16:54:20 https://kazuhikoarase.github.io/simcirjs/ Mar 19 16:55:15 I need to update the melange proposal with the deployment part Mar 19 16:55:18 It would certainly do for GPIOs to start. Mar 19 16:55:58 Any of the more complex buses, like I2C and SPI, would really depend on what you were talking to on the other end, anyway. Mar 19 16:56:14 GPIOs with digital I/O only would be the scope Mar 19 16:56:17 But you could take an approach with generic components like those for GPIO with no trouble. Mar 19 16:56:37 OK. That is quite reasonable. Mar 19 16:57:10 I am currently developing a very simple proof of concept Mar 19 16:57:54 I would suggest that you expand your current proposal to put in as much detail as possible. In particular, adding in a list of exactly what you are proposing to deliver at the end of the project under "Goals". Mar 19 16:58:49 Individual items such as "simulated GPIO behavior", "Side-by-side use with bonescript", etc. Mar 19 16:59:00 Don't leave anything out. Tell us everything. Mar 19 16:59:08 hendersa: will do. Thanks for the help. Mar 19 16:59:22 anything else? Mar 19 16:59:56 Start with that and we'll take it from there. If you put in a more detailed proposal, you'll begin to get more comments from the mentors through the Melange system. Mar 19 17:01:10 I will certainly do that, thank you for your help Mar 19 17:01:31 Good luck! Mar 19 17:02:20 Well, that is all the time I can spare for now. Good luck on your proposals, everyone. Get those first round of proposals in and get some feedback. Mar 19 17:11:14 * nerdboy present but mostly unconscious Mar 19 17:14:23 that should take about a minute... Mar 19 17:15:09 ketul: please take off the /msg... this is a discussion after all... Mar 19 17:15:32 hi nerdboy Mar 19 17:15:40 I was reading the project idea about Orbital Imaging Cubesat Mar 19 17:16:23 The ultimate goal is a small cubesat-based orbital imaging platform ? can you give me any realistic example for that. I understood that it is all about data acquisition. just want to know that what will be final output? Mar 19 17:16:50 actual form-factor (1U, 2U, etc) is future, i would say initial goal is bench-test detection and generating acquisition data Mar 19 17:17:36 best sensor suite is not fully defined either Mar 19 17:18:12 something with IMU/accelerometer/gyroscope plus the magnetometer Mar 19 17:18:28 nerdboy: can you give me any reference link about the goal. Mar 19 17:18:50 * nerdboy assumes jkridner would cough up a few bucks worth of sensors for this one Mar 19 17:19:15 I have some experience with interfacing of different sensors with BBB Mar 19 17:20:29 nerdboy: as I do have background of Electronics I do meetup all the hardware skills requirements. Mar 19 17:22:51 https://www.sparkfun.com/products/12636 <= several baords like this one Mar 19 17:24:16 reference link for actual satellite? should be a few out there... Mar 19 17:25:34 nerdboy: ok got the point Mar 19 17:26:09 it's pretty much all real-time data acquisition/processing... keeping track of your own orientation calculate pointing vector between you and detected boject Mar 19 17:26:18 *object even Mar 19 17:27:26 open questions include data rates, detector sensitivity, etc Mar 19 17:28:28 +/- 2 gauss might work depending on resolution Mar 19 17:30:31 nerdboy: ok. then which OS will you suggest for that ? So that I can get started doing some stuffs with my BBB. Mar 19 17:31:35 shouldn't matter too much Mar 19 17:32:03 same gpio interface Mar 19 17:32:37 breakout boards are generic enough Mar 19 17:32:45 so debian will work fine right? Mar 19 17:32:55 as long as SPI/i2c is good Mar 19 17:33:02 should be Mar 19 17:33:43 ok so I should first work on SPI and i2c through BBB . I might get some available projects Mar 19 17:34:07 spi would be fastest Mar 19 17:34:32 yep that will. Mar 19 17:34:39 well, vs. i2c stuff Mar 19 17:35:16 yea Mar 19 17:36:28 the not-too-crappy boards/sensors typically have both Mar 19 17:37:03 what datarates are you targeting? Mar 19 17:37:30 as of now actually I do have only BBB as a hardware. No sensors Mar 19 17:37:45 good question Mar 19 17:38:51 not even sure what "fast enough" is... Mar 19 17:39:39 * nerdboy has mostly weather sensor stuff Mar 19 17:40:14 might have a spi/accelerometer i can test... Mar 19 17:40:39 https://www.sparkfun.com/products/12636 has max 100MHz spi clock freq Mar 19 17:41:37 what are you sensing? Mar 19 17:41:51 earthquakes? cars? Mar 19 17:42:26 these sensors are pretty trivial to work with Mar 19 17:43:16 sry it is 10Mhz Mar 19 17:44:25 for a lot of things spi is overkill Mar 19 17:51:55 ds2: this is related to nano satellite I think Mar 19 17:51:58 http://en.wikipedia.org/wiki/CubeSat Mar 19 17:52:21 alexanderhiam: the python firmata which I am making is not actually firmata protocol per se. what do you want me to mention about it in the proposal ? Mar 19 17:56:06 nerdboy: you there? Mar 19 17:57:39 looking at hardware... Mar 19 17:57:59 ok Mar 19 17:58:07 :) Mar 19 17:58:33 maybe separate magnetomoeter/other for high speed data Mar 19 17:59:01 internal stuff could use i2c Mar 19 18:01:58 so imu is for the positioning of cubesat.? Mar 19 18:04:15 nerdboy: am I right? Mar 19 18:07:22 well, main /whatever/ is internal position/orientation Mar 19 18:07:44 different boards have different sensors Mar 19 18:08:18 https://www.tindie.com/products/ozzmaker/berryimu-an-accelerometer-gyroscope-magnetometer-and-barometricaltitude-sensor/ <= another one Mar 19 18:10:11 +1 Mar 19 18:10:26 hmm, probably not rated for space... Mar 19 18:10:28 ketul: what are you measuring with those sensors? Mar 19 18:10:52 rocket acceleration? Mar 19 18:11:35 ds2: position of the satellite. You can think like the IMU for quadcopter Mar 19 18:12:18 quadrortor will take all the decisions based on imu data Mar 19 18:13:26 like that so all the data acquisition of satellite would be done using this kind of sensors. Mar 19 18:16:31 * nerdboy thinking phase 1 should be "get max data to prus and process" Mar 19 18:17:28 nerdboy: and we need to find out with rating of space. right? Mar 19 18:18:44 ds2: am I ..? Mar 19 18:19:51 <_av500_> ds2: there was a student interested in the ADC project earlier Mar 19 18:21:47 ketul: doesn't a sat move a lot faster then a quad? Mar 19 18:22:20 _av500_: roughly what time? (trying to narrow down the logs) Mar 19 18:22:41 dmitescu i think is his nick Mar 19 18:22:47 got it Mar 19 18:22:54 ds2: ya i know but that was the example Mar 19 18:23:19 ketul: yes... but my point is this - the faster it goes, the faster you need samples Mar 19 18:23:30 (nyquist and all the other good stuff) Mar 19 18:23:47 <_av500_> 13:34 < dmitescu > Good day. I am a Computer Science Undergrad. Student in Germany, and I am interested in ADC improvement in Beagle Boards. Mar 19 18:23:51 I seem to be coming on at all the wrong times thesedays Mar 19 18:24:05 _av500_: thanks. I found it once vvu gave me the nick Mar 19 18:24:57 I'll see what I can do to shuffle the time around Mar 19 18:25:38 I do wonder - do cubesats require EAR/ITAR paperwork Mar 19 18:25:58 maybe question is what is best data rate from each possible interface... Mar 19 18:26:50 nerdboy: I'd think the sensor itself might not be able to sample fast enough Mar 19 18:29:17 what is the required data rate for cubesat? Mar 19 18:29:33 alexanderhiam_\: the python firmata which I am making is not actually firmata protocol per se. what do you want me to mention about it in the proposal ? Mar 19 18:29:42 alexanderhiam_: the python firmata which I am making is not actually firmata protocol per se. what do you want me to mention about it in the proposal ? Mar 19 18:30:34 nerdboy: may be any ongoing satellite's data can we get? Mar 19 18:30:51 kiran4399: well... don't call it firmata if it's not firmata Mar 19 18:30:56 not sure Mar 19 18:31:21 hopefully university stuff? Mar 19 18:32:02 yep. we need to ask some space org. first Mar 19 18:32:03 alexanderhiam_: do you have some sort of technical name or something like that which I can put it in proposal ? Mar 19 18:33:12 kiran4399: well it's a remote procedure call framework. Firmata is the name of one specific rpc framework Mar 19 18:36:26 alexanderhiam_: so we can call python rpc framework ?? Mar 19 18:37:36 rather then IMU, I wonder if you are better off doing a SDR GPS Mar 19 18:38:09 * nerdboy was also wondering if gps would work Mar 19 18:39:33 kiran4399: sure, if it fits that description. You're designing it, so you'd know better what it is then I. The proposal is the place to describe to us what your design is, just don't call it firmata if it's not implementing firmata Mar 19 18:41:36 alexanderhiam_: yes sure... one more thing... I hope you know that my project is divided into 3 parts.. Mar 19 18:42:26 alexanderhiam_: devoloping pruspeak.. adding python rpc_framework.. and adding support for node.js Mar 19 18:44:00 so it is like a sort of bottom-up approach I am following.. Mar 19 18:44:10 finish off pruspeak development Mar 19 18:44:21 write the rpc... Mar 19 18:44:47 and then go to node.js.. Mar 19 18:46:38 be in and out... nutty day Mar 19 18:47:15 nerdboy: then what are we supposed to do? Mar 19 18:48:04 ask jkridner for gps hardware? Mar 19 18:49:23 https://www.sparkfun.com/products/10920 <= like that one Mar 19 18:49:45 alexanderhiam_: one more thing... I was referring to botspeak for arduino...I saw commands like tmr and servo .. can they be implemented for pruspeak ? Mar 19 18:50:10 kiran4399: that would be great Mar 19 18:50:18 Yes... Mar 19 18:51:15 nerdboy: didn't get what will be role of gps? Mar 19 18:57:57 alexanderhiam_: I am trying to enable the gpio's with the starterware from pru... i am doing fine until I got stuck in a problem... is it that the resource are supposed to be freed up before using them ? Mar 19 18:58:53 which resources? Mar 19 18:59:05 nerdboy: read http://www.nasa.gov/content/goddard/cubesat-instruments-to-demonstrate-nasa-firsts/#.VQscAI4c6hs Mar 19 18:59:58 alexanderhiam_: gpio pins.. I think.. Mar 19 19:01:26 I've haven't used starterware, but if it's reading/writing te GPIO registers directly it shouldn't get hung up as long as you haven't changed the base DT which enables the GPIO module clocks Mar 19 19:01:48 alexanderhiam_: I mean the pins needs to be released from the ARM cortex A8 before taking them to PRU right ?? Mar 19 19:02:57 no, the only 'control' model is the exporting, which is just part of the kernel driver to let userspace programs request that pins be left alone by the driver Mar 19 19:05:03 well, that's assuming the PRU OCP port is enabled Mar 19 19:05:34 what is DT ? Mar 19 19:05:39 Device Tree Mar 19 19:05:53 as expected. Mar 19 19:06:45 but in the dtb I saw only some pins were enabled... I mean the one written by panto.. Mar 19 19:07:57 nerdboy : ping Mar 19 19:08:04 you're thinking of the pinmux settings. That's a separate hardware block that the GPIO signals go to from the GPIO modules that sets which signals go to the AM335x's physical pins Mar 19 19:08:35 The GPIO module's know nothing of the existence of the multiplexers Mar 19 19:09:00 kiran4399: have you tried doing GPIO without starterware? Mar 19 19:09:13 doing the memory access yourself that is Mar 19 19:10:11 alexanderhiam_: yes... It was a simple register access.. Mar 19 19:10:13 https://github.com/kriswiner/MPU-6050/wiki/Affordable-9-DoF-Sensor-Fusion <= some actual data Mar 19 19:10:58 ketul: gps would be alternative (or even in addition to) to imu sensors Mar 19 19:11:05 kiran4399: so if starterware isn't working it must be doing something differently, have you looked through the source? Mar 19 19:11:30 for locating the exact position .? Mar 19 19:12:03 alexanderhiam_: Yes... I don't know why I am unable to do it with pru.. I went through the drivers.. It is not so complicated... just uses simple API's.. Mar 19 19:12:30 I meant have you done GPIO from the pru Mar 19 19:13:20 alexanderhiam_: yes... from pru only... Mar 19 19:13:35 so what's starterware doing differently? Mar 19 19:13:55 alexanderhiam_: I am unable to do gpio access from pru with starterware !!! :-) Mar 19 19:15:43 nerdboy : hey. Beside analysing the real time capability of different kernels in the beagleRT project what are the other goals ? Mar 19 19:15:51 alexanderhiam_: Using starterware it is very simple.. we just need to set the OCP... enable the gpio module... reset the module....set the pin direction.... write it..... Mar 19 19:17:07 alexanderhiam_: But all I get is some wierd problem... when I run the pru with modified firmware.... and run a botspeak command.... user led's 2 and 4 go solid... no output comes after running the command.. Mar 19 19:17:14 kiran4399: why are you resetting the module? The kernel is already enabling the GPIO modules when you boot, if you then reset them you could screw up anything using any other GPIO pins Mar 19 19:17:45 alexanderhiam_: OK... but without that also I am facing this issue... Mar 19 19:18:23 alexanderhiam_: I don't know why 2 user led's are suddenly going solid without any progress.. Mar 19 19:18:26 alexanderhiam: not sure there is one, as i didn't write it... Mar 19 19:18:32 well you've got the sterware source, look at what it's doing Mar 19 19:18:39 er, altairpearl Mar 19 19:19:40 hey Mar 19 19:19:49 alexanderhiam_: OK... Mar 19 19:19:52 charcterize and integrate best one maybe? Mar 19 19:20:15 hmm . Mar 19 19:20:30 CONFIG_RT_GROUP_SCHED is default in mainline Mar 19 19:20:50 alexanderhiam_: just for quick grasp.. what does tmr command do ?? Mar 19 19:21:15 er, RT_GROUP_SCHED is current mainline feature Mar 19 19:21:38 In the group mail i came to know that this was not enough and additional bug fixes for the RTand xenomai kernel would be preferred Mar 19 19:21:55 if there's really no difference that would be good to know Mar 19 19:22:44 kiran4399: something timer related perhaps? Mar 19 19:22:44 hmm Mar 19 19:22:56 where's your code? Mar 19 19:22:59 in theory RT_GROUP_SCHED should get close Mar 19 19:23:19 7 msec audio latency maybe? Mar 19 19:23:43 * nerdboy has no recent data Mar 19 19:24:08 oooh Mar 19 19:25:18 i would guess actual testing/data analysis is the goal Mar 19 19:25:45 nerbboy: do we need any camera ? Mar 19 19:25:45 for imaging Mar 19 19:25:50 nerdboy : using the stand test methods mentioned in that research paper ? those 6 tests ? Mar 19 19:26:05 what research paper? Mar 19 19:26:30 https://www.osadl.org/fileadmin/dam/rtlws/12/Brown.pdf ???? Mar 19 19:26:37 oh, the bronw one? Mar 19 19:26:45 yea Mar 19 19:27:22 depends on creative your tester hat is i guess Mar 19 19:27:26 *how Mar 19 19:27:47 hmmm ok Mar 19 19:28:21 need to scope it for gsoc Mar 19 19:28:59 yea I'm trying to decide between these two projects beagleRT and cubesat :3 Mar 19 19:29:28 i want to work with cubesat after i learnt more about the PRU Mar 19 19:30:14 ahh. Mar 19 19:30:26 right now I have a 3DR ublox gps module with compass Mar 19 19:32:12 looks like they used a 2.6 kernel? Mar 19 19:32:48 so replicating their test setup would be a good comparison Mar 19 19:34:13 nerdboy: what about the cubesat? what kind of stuff should I include in my proposal? Mar 19 19:34:36 any camera do we require for the same? Mar 19 19:35:31 camera might be a bit much for short project Mar 19 19:36:15 could just use a webcam Mar 19 19:36:24 trying that approach with 2 magnetometers might be good pru task Mar 19 19:37:29 use "IMU" magnetometer data for internal measurements, and second one for "external" Mar 19 19:39:03 this part should really be a data processing task Mar 19 19:39:14 gps Mar 19 19:39:30 nerdboy: ok Mar 19 19:40:21 sorry accidental typing!! Mar 19 19:40:38 probably end up with a gps too... Mar 19 19:40:40 nerdboy: didn't get the 'internal' Mar 19 19:41:11 take one of those 9/10 DOF breakout boards Mar 19 19:41:51 the gyro/accel can get data from aux input direct from magnetometer Mar 19 19:42:20 depending on board, can also read directly from magnetometer Mar 19 19:43:03 subtract "noise" measurement from that one from what you measure from a second magnetometer Mar 19 19:43:46 second one could presumably be deployed on a little pop-out mast or whatever Mar 19 19:48:37 need good data for pointing solution, then point camera... Mar 19 19:49:41 nerdboy: so our final goal would be data processing through magnetometers that may be helpful for cubesat in future. Mar 19 19:50:00 might even take a flock of these things to do anything really useful... Mar 19 19:52:28 nerdboy: much of the work done in project Interacto. So that would be our baseline. Then we will do implement nice data processing Mar 19 19:53:03 did you find an upstream for that? Mar 19 19:53:33 nope Mar 19 19:53:33 i got https://github.com/cclark2/interacto_bbone_cape Mar 19 19:54:17 last updated in 2013 Mar 19 19:54:51 software is sensor/interface specific anyway Mar 19 19:55:06 should work with same sensors over i2c Mar 19 19:56:11 should be the same with sparkfun or adafruit board with same sensors Mar 19 19:56:45 * nerdboy will update wiki page today Mar 19 19:58:34 thanks nerdboy. It would be helpful for making effective proposal. Mar 19 20:01:23 alexanderhiam_, alexanderhiam: Should the content in my proposal be as if answering to the questions given in the template or is it that the content itself has to contain the answers ? Mar 19 20:21:47 Is user ds2 here? Mar 19 20:28:25 ??? Mar 19 20:28:57 dmitescu: yes? Mar 19 20:35:56 Okay. Mar 19 20:36:18 I am interested in enhacing the ADC driver for BeagleBoard. Mar 19 20:36:35 dmitescu: what do u know about it? Mar 19 20:36:49 I did the required task for the proposal (which I will submit in a few minutes). The next step is to plundge into the source code and schematics and search for solutions. Mar 19 20:37:07 dmitescu: what is ur local time? i am in and out Mar 19 20:37:27 At the moment it's 21:37. GMT+2. Mar 19 20:38:01 dmitescu: good... did u go through the docs (trm mainly)? Mar 19 20:38:36 I don't know much about it - yet. As I said I will look soon into schematics and code. But as far as I've read, the ADC sampling will have to be done by the ADC software, which is obviously better Mar 19 20:38:37 No. Mar 19 20:38:39 dmitescu: ok.... r u a night person? I need to pop off in about 2 mins but will be back...i am in and out so if u hang around... Mar 19 20:38:41 'trm'? Mar 19 20:38:48 Yes, I can wait. Mar 19 20:39:08 technical reference manual Mar 19 20:39:12 dmitescu: send mail to the bb gsoc list if i am not around. Mar 19 20:39:47 Alright. Mar 19 20:39:50 sampling in hw is imo more important. but if u find requests , do mention them in ur proposal. Mar 19 20:40:12 dmitescu: or just ask away. i will eventually see it. Mar 19 20:40:50 problem is the current driver sucks. and people are doing all sorts of no no to bypass them. Mar 19 20:41:23 dmitescu: one thing that needs to be kept in mind is we cannot break the tsc which uses the same hw. Mar 19 20:41:40 sorry, gotta go. back sometime later. Mar 19 20:42:06 Okay, thank you. Mar 19 21:01:25 hi Mar 19 21:01:40 i developed a proof of concept for my project idea Mar 19 21:01:40 ktrajkovski.com/bonesim/sample.html Mar 19 21:01:47 http://ktrajkovski.com/bonesim/sample.html Mar 19 21:47:34 back Mar 19 21:49:23 Alright. Mar 19 21:49:45 In my proposal, should a sketch for the solution be presented? Mar 19 21:50:37 whatever u feel comfortable with Mar 19 21:51:11 it needs to address communityneeds (i.e. BB ML posts) and at the same time fit in the gsoc time frame Mar 19 21:51:45 if u can convince folks of that, it would be a good start Mar 19 21:52:21 Alright. I'll post the ARM program on the mailing list (as it is requested on the proposal), and in the meanwhile, I'll start working and see what I can do. Mar 19 21:52:24 Thank you for you support Mar 19 21:52:30 iirc, some of the driver is still not completely IIO compliant Mar 19 21:52:58 so making that could be a good thing but donno if anyone needs that now. Mar 19 21:53:20 just ask.... folks will try to answer even if i am not around. Mar 19 21:53:26 Okay. :D Mar 19 21:59:09 any love for my circuit simulator? :) http://ktrajkovski.com/bonesim/sample.html Mar 19 22:01:45 explain what is it good for? Mar 19 22:07:19 Demonstrates how a beaglebone simulator project may work Mar 19 22:07:30 https://groups.google.com/forum/#!topic/beagleboard-gsoc/CC2CZ6pmuV4 Mar 19 22:08:10 The BBB component is a Beagle Bone you can program with bonescript Mar 19 22:08:54 currently it is just a proof of concept Mar 19 22:16:44 I am on the list Mar 19 22:16:57 i don't see the value this would add Mar 19 22:17:36 for example - how is this different from running SPICE on the BBB besides this being digital only Mar 19 22:20:35 This should be for new users trying to get up and running with simple projects Mar 19 22:21:01 and users that don't own a Beagle board for example Mar 19 22:21:17 okay... so is this like a QEMU powered Beagle? Mar 19 22:22:10 similar to that Mar 19 22:22:18 but it only supports bonescript Mar 19 22:22:41 and everything IO is sumulated in javascript Mar 19 22:23:48 how complex of a circuit are you simulating? Mar 19 22:24:29 right now it just demonstrates how I/O pins can be programmed with bonescript Mar 19 22:25:30 but at the end of GSoC, what will it do? Mar 19 22:27:36 It will offer full simulation of digital components, will allow for full configuration of pin modes for I/O, it will have a pretty code editor, it will have all bone101 samples ported Mar 19 22:28:07 what kind of digital components? Mar 19 22:28:27 i don't think my questions are that far off but you seem to be avoiding answering anything specific. Mar 19 22:29:42 sorry, what i meant was Mar 19 22:30:53 it will allow users to do everything they can currently do with bonescript and a beagle bone black's GPIO pins Mar 19 22:31:25 okay - so you are simulating a subset of the BBB? Mar 19 22:31:38 exactly Mar 19 22:31:50 then why not build on top of qemu? Mar 19 22:31:56 why this branch off into JS land? Mar 19 22:32:08 Why would i build on top of qemu? Mar 19 22:32:23 Because this is much more accessible Mar 19 22:32:24 because folks other then beginners can get value from it Mar 19 22:32:40 Everything is going on in the browser, so it's easy for beginners Mar 19 22:32:47 qemu does the cpu work, you largely need to add the gpio + pinmux peripherals to it Mar 19 22:33:46 beginners are using bonescript anyways, all the samples in bone101 are bonescript Mar 19 22:35:43 How would the circuit simulator could be coupled with qemu? Mar 19 22:36:23 you implemented as a peripheral Mar 19 22:36:29 qemu emulates the SoC Mar 19 22:36:45 then the user just fires up JS and things just work. Mar 19 22:39:11 vvu, praveendath92: I've written my proposal and still working on it. But would like to get some feedback from you. Can you take a look? Mar 19 22:39:24 What would qemu add as a value versus doing this? Mar 19 22:39:26 Sure Mar 19 22:39:53 qemu simulates other things Mar 19 22:40:07 if done right, qemu would let you boot a stock image Mar 19 22:40:12 Qemu first needs to be installed, BBB image should be installed also Mar 19 22:40:18 https://azizulhakim2002.wordpress.com/2015/03/19/gsoc-2015-proposal-for-beagleboard-android-remote-display/ Mar 19 22:40:23 this is my link Mar 19 22:40:54 ou would run QEMU on a desktop/laptop Mar 19 22:41:07 you Mar 19 22:41:45 The point of the project is to provide a multiplatform solution that runs in the browser and lets users experiment with simple circuits and bonescript Mar 19 22:42:01 without needing to install anything Mar 19 22:42:27 *shrug* it is your proposal Mar 19 22:42:37 i don't see the point of doing things that way but that's just me. Mar 19 22:43:10 i do not see what qemu adds as a value over this Mar 19 22:43:42 do you have potential mentors? Mar 19 22:43:46 i did find this when researching what to do http://bellard.org/jslinux/tech.html Mar 19 22:44:01 not really Mar 19 22:44:47 could you suggest someone? Mar 19 22:45:17 not really Mar 19 22:45:24 i don't see any long term value to it Mar 19 22:45:42 if all you'll get done is GPIO, what will happen to this after GSoC? Mar 19 22:46:03 there is a lot more to the I/O then GPIO -I2C,SPI,ethernet,video,etc Mar 19 22:46:32 It can be used to demonstrate how the basic examples really work Mar 19 22:46:46 OTH - convincing me to be interested in JS is a very very tall order ;) Mar 19 22:47:12 Instead of this http://diegotc.github.io/bone101/Support/GSOC/views/tutorial.html?gistid=855e0d8d0b6f6f53f06b#2 Mar 19 22:47:40 if you have gotten that far, just get a board. Mar 19 22:47:41 There will be an actually simulated circuit Mar 19 22:48:20 now if you were doing a virtual hook up of the BB to say verilog/vhdl or schematic equiv.... that would be a lot more interesting ;) Mar 19 22:49:31 I'm not too much a hardware guy Mar 19 22:50:44 one thing you will need to do is locate potential mentors Mar 19 22:50:53 So i really couldn't be of help there Mar 19 22:51:13 verilog/vhdl is "software" til you do the magic on it Mar 19 22:51:54 But it's used for hardware, so it does not interest me much Mar 19 22:52:16 but so are gpios :D Mar 19 22:52:46 but in this context, it would just be simulated with javascript Mar 19 22:52:58 and there's not much philosophy to gpios Mar 19 22:52:59 so would the VHDL/Verilog ;) Mar 19 22:53:08 who is Julian Duque? Mar 19 22:53:47 julianduque: could you be my mentor Mar 19 22:54:32 ds2: Yes, but i have no idea how those work Mar 19 22:54:41 there are several variants of that stuff... Mar 19 22:54:58 plus some actual tools for verification Mar 19 22:56:09 with all these talk about the PRU - has anyone gotten PRUdebug to just work? Mar 19 22:56:17 http://www.cprover.org/ebmc/ <= example that does VHDL Mar 19 22:57:04 look at the icarus stuff Mar 19 22:57:04 daniel was taking PhD students not that long ago... Mar 19 22:57:44 might be able to map the icarus simulation stuff to qemu's virtual GPIO Mar 19 22:57:55 http://www.cs.cmu.edu/~modelcheck/vcegar/ <= this one does Verilog Mar 19 22:57:57 then you can flag in the UI if there is a electrical conflict Mar 19 22:58:21 cool stuff... Mar 19 22:59:11 * nerdboy pimping for formal verification Mar 19 23:00:04 brain still addled with some kind of andromeda strain, but hey... Mar 19 23:00:18 Those don't run on the web Mar 19 23:00:25 :) Mar 19 23:00:48 and i have zero idea what they are and how they are used Mar 19 23:03:53 I have tried doing some simple circuits before on a breadboard Mar 19 23:04:14 but nothing hardcore that would require verilog or vhdl Mar 19 23:54:24 azizulhakim: will take a look :) Mar 19 23:54:33 but looks ok on a small glance Mar 19 23:58:39 okay. thanks Mar 20 00:01:36 av500: my initial proposal for android remote display project https://azizulhakim2002.wordpress.com/2015/03/19/gsoc-2015-proposal-for-beagleboard-android-remote-display/ Mar 20 00:02:01 would love to get some feedback if possible Mar 20 00:08:21 Hello Mar 20 00:08:38 i'm interested in the Android under Angstrom GSoC project Mar 20 00:08:55 *crickets chirping* Mar 20 00:10:43 do i get that correctly that android and angstrom would be running paralel? Mar 20 00:12:09 sort of Mar 20 00:12:15 problem with that project is android changes too fast Mar 20 00:12:33 I know that :D Mar 20 00:12:35 azizulhakim: please do a bit more investigation how to include everything into one driver and check if you can match devices automatically, without the user to add vid/pid Mar 20 00:12:45 what version of android are we talking about? Mar 20 00:13:10 that project might have been proposed in the GB time frame! Mar 20 00:13:55 so is it still actual? Mar 20 00:14:06 that was long ago.. Mar 20 00:15:00 * nerdboy just noticed cricket boy's name is on it... Mar 20 00:15:16 yes very very long ago Mar 20 00:16:47 so the latest android version for beagleboard is gb? Mar 20 00:16:53 no Mar 20 00:17:00 might be a KK floating around now Mar 20 00:17:31 for beagleBOARD? or beagleBONE? Mar 20 00:18:20 which is for the project? Mar 20 00:19:24 it could go either way Mar 20 00:20:30 is this kk you talked about stable enough to use for this project? Mar 20 00:29:14 probally Mar 20 00:29:29 there is another guy here, hensomething that'd know more **** ENDING LOGGING AT Fri Mar 20 02:59:59 2015