**** BEGIN LOGGING AT Wed Jul 08 02:59:58 2015 Jul 08 03:39:55 morning Jul 08 03:46:26 Gm Jul 08 04:02:26 * neemo has that strange feeling of exasperation which you can only get by receiving 8 kb/s from GitHub... on a 200Mb connection Jul 08 04:02:43 (sigh) Jul 08 04:02:55 alexanderhiam: awake, per chance? Jul 08 05:50:20 neemo: do you have something to push? Jul 08 08:38:11 good morning folks :-) Jul 08 08:52:51 good morning! Jul 08 09:41:25 shubhangi: ping Jul 08 09:57:26 nerdboy: nothing to write home about Jul 08 09:58:50 nerdboy: I started making an overall design how to put all the pieces together and figured half way through that it's crappy and that I don't know enough about sysfs and LKMs Jul 08 09:59:56 went through all of the material in http://derekmolloy.ie/kernel-gpio-programming-buttons-and-leds/ and a load of other places and I think I got the hang of it now Jul 08 10:01:57 now I'm working on the overview, writing requirements and what the API should look like Jul 08 10:02:47 neemo maybe you know about kernel newbies website. they have a lot of information. there might be something interesting. Jul 08 10:05:53 arianepaola_: do you have a link Ariane? Jul 08 10:07:04 neemo it is kernelnewbies.org Jul 08 10:08:24 thanks arianepaola_ , I'll have a look :) Jul 08 10:09:44 also, for LKM and GPIO stuff, derek's blog above was extremely useful, especially since it's based on the BBB Jul 08 10:10:05 so can't recommend that enough Jul 08 12:15:21 Abhishek_: pong Jul 08 13:19:57 * Abhishek_ checks in Jul 08 13:20:11 Abhishek_: ping Jul 08 13:20:42 Hi shubhangi Jul 08 13:20:50 Hey Jul 08 13:21:43 Hi jkridner Jul 08 13:21:56 my local source tree is all messed up :( trying to sort it out to get the patches done Jul 08 13:22:23 a couple of more files needed some changes .. apart from your patches Jul 08 13:22:31 then all worked well Jul 08 13:22:39 jkridner: I now have an IPS panel 1024*600 for my display cape, very nice viewing angles. Jul 08 13:23:20 shubhangi: good. Are you using git-format patch@ Jul 08 13:23:24 *? Jul 08 13:24:06 yes. Jul 08 13:32:33 Abhishek_: cool. Jul 08 13:33:01 shubhangi: your local source tree wildly different than your git tree? Jul 08 13:33:26 shubhangi: got some work you are trying to save instead of going back to the last known good point in your tree? Jul 08 13:35:13 jkridner: yes. i checked checkout out to a previous good state. all seems fine Jul 08 13:36:09 hurray for version control! Jul 08 13:36:10 checked out* Jul 08 13:36:48 Nice Jul 08 13:37:12 Now just waiting to catch rcn-ee ;) Jul 08 13:38:04 Abhishek_: apparently the lines which were commented ( which gave the reset configuration missing ) were the problem Jul 08 13:38:21 shubhangi: You can rebase your branch and reorder commits as well Jul 08 13:39:15 shubhangi: Their being present was the problem or their being commented out Jul 08 13:39:22 ? Jul 08 13:39:36 being commented out Jul 08 13:39:51 later during probe it deasserts pruss reset Jul 08 13:40:02 ie pull it out of reset Jul 08 13:40:31 * shubhangi rebasing and reordering sounds like advanced git :? Jul 08 13:41:05 But then why was it not working previously and had to be commented out? Jul 08 13:41:33 it had to be present .. not commented out Jul 08 13:42:00 In my case, their being present was causing the probe to fail Jul 08 13:42:03 pruss_remoteproc from 3.14 as it is .. no changes needed in that Jul 08 13:42:19 yes . that happended for me too Jul 08 13:42:43 it needs a differnet file to be patched .. sending link Jul 08 13:43:31 Abhishek_: https://github.com/beagleboard/linux/blob/4.1/arch/arm/mach-omap2/pdata-quirks.c Jul 08 13:43:55 this file provides that platform data for pruss when the platform device is created Jul 08 13:44:19 which this does not in case of 4.1 ... in 3.14 it did Jul 08 13:51:14 shubhangi: I see. When do you plan to push it? Jul 08 13:51:21 Abhishek_: i can try moving this to the driver code itself .. because the way it is right now the driver would break each time the kernel version upgrades (unless changes are pushed to mainline). Jul 08 13:52:26 or maybe theres a reason as to why things are the way they currently are. Jul 08 13:52:30 Could you make a separate branch so that I could see the changes you made so far? Jul 08 13:53:28 Don't push it to master yet. Jul 08 13:57:36 Abhishek_: doing it right now Jul 08 14:00:22 shubhangi: Do you know about git send-email? Jul 08 14:01:12 i read about it while reading on how to prepare patches. yet to try Jul 08 14:01:22 i'll do it. Jul 08 14:01:44 gimme an hour or two Jul 08 14:42:01 neemo: where's your first attempt at the API? Jul 08 14:45:36 neemo: I'm confused as to why you didn't push anything. If it didn't seem right then you should have pushed it and asked nerdboy and I to take a look, that's what we're here for! Jul 08 14:46:35 instead we're another week in and we still haven't seen anything relating to the framework, which is not a very good sign from our perspective Jul 08 15:07:11 mornin Jul 08 15:21:10 moin Jul 08 15:22:25 40mins to meeting Jul 08 15:22:48 Abhishek_: you email id to send patches to ? Jul 08 15:31:08 alexanderhiam: my first attempt got sent to the trash bin, since it was utter crap or so it seemed Jul 08 15:31:43 I tried fixing it to send it in, but I was quite frankly ashamed of how crappy it was Jul 08 15:32:16 I've since went on to read all the useful stuff I've bumped into on LKMs and sysfs Jul 08 15:32:47 neemo: what was it? All we are asking for for a first step is prototype function declaration for the C API. No kernel stuff involved Jul 08 15:32:52 and now I'm working based on http://www4.in.tum.de/~blanchet/api-design.pdf on the api Jul 08 15:33:14 I think you're WAY overthinking it Jul 08 15:33:44 alexanderhiam: yeah I've figured that, but couldn't get anything that makes sense without figuring how the overall system would fit together Jul 08 15:34:06 I think I understand now and I'll hang out a bit later if you're guys up for it to make a rundown of the stuff I had in mind Jul 08 15:34:13 just things like `float BeagleSat_getRotation();` Jul 08 15:34:44 yeah, think I understand that now Jul 08 15:34:58 but was mighty confused 5 days ago Jul 08 15:35:11 that's the kind of thing where you really need to be asking us if you don't understand what we're asking for Jul 08 15:35:14 that's why you speak up Jul 08 15:35:53 from our perspective we asked for something small a week ago and then never heard anything from you Jul 08 15:36:00 as in "WTF am I doing?!? Anybody?!?" Jul 08 15:36:23 even if we're not around there's plenty of other people here that could help with that stuff Jul 08 15:36:30 alexanderhiam: nerdboy my apologies for not pinging you on that stuff, was confused since I lacked a lot on the API and the overall system side Jul 08 15:36:59 so I went into the only logical thing I could think of, read a shitton of stuff so I get a better idea of how I'm supposed to do this Jul 08 15:37:21 it's just a user-space api right now... like the doxygen example header... Jul 08 15:37:42 I covered all the sysfs and MKL stuff pretty good and I'm comfortable with that (should have maybe done that a bit later) Jul 08 15:37:49 and read the API guide and a couple of other Jul 08 15:37:54 maybe alternate reading and doing Jul 08 15:38:02 now I've gotaa requirements list and some usecases Jul 08 15:38:07 so let's use this time before the meeting to make sure we're on the same page Jul 08 15:38:17 "try it and see..." then go read more if it looks wonky Jul 08 15:38:31 and will start moving from that list into API code Jul 08 15:38:56 hard to yell at a guy for wanting some requirements... Jul 08 15:39:12 but you do need to speak up more often Jul 08 15:39:29 do you understand what file we're asking for and where it fits in the whole system? Jul 08 15:39:35 nerdboy: yeah, more of that I guess, was a bit hard to try without an idea of what/how/where :/ Jul 08 15:40:01 you can compile a header file ya know... Jul 08 15:40:02 alexanderhiam: you're asking for the API header, the general outline of the API, e.i. how an enduser will most easily interact with the systems Jul 08 15:40:34 in this case, first point in building a C API Jul 08 15:40:43 am I on track? Jul 08 15:41:30 ^/same page/ Jul 08 15:41:31 right, more specifically the header file that gets included in an end user's program and defines a set of function which abstract away the interactions with the kernel driver Jul 08 15:41:47 ok, got that Jul 08 15:42:10 an that set of functions should interact with the kernel driver through sysfs entries Jul 08 15:42:52 so that a more advanced user has the flexibility to interact with the sysfs entries more directly if he so chooses/or upgrade the API with new functions Jul 08 15:42:55 i would look at libsysfs interface rather than poke at /sys Jul 08 15:44:06 with something like this you might take an object oriented approach, where you have a `struct BeagleSat` that holds information like what algorithm to use and what sensors are attached, then functions named `BeagleSat_xxx` that get passed a pointer to a BeagleSat struct Jul 08 15:44:44 but that is implementation so flip back to "user" api interface Jul 08 15:45:04 hey rohith Jul 08 15:45:12 any update? Jul 08 15:45:23 Vmayoral|pc: hi Jul 08 15:45:34 ankur: nice post on the i2c app Jul 08 15:45:55 anujdeshpande: thanks Jul 08 15:46:05 vmayoral: Getting errors while compiling. Have been troubleshooting? Jul 08 15:46:12 ankur: love the screenshot and the function desc Jul 08 15:46:16 nerdboy: I'm reading some not so good things about libsysfs Jul 08 15:46:20 s/?/. Jul 08 15:46:34 alexanderhiam: such as? Jul 08 15:46:56 anujdeshpande: any improvement you would you would like to suggest? Jul 08 15:46:57 not actively maintained, makes incorrect assumptions about sysfs Jul 08 15:47:01 since we are coming towards the end, you could spend some time on doing similar stuff for the rest of the apps too. probably on the github wiki of your repo Jul 08 15:47:12 shubhangi: The patches look good to me, you can go ahead and mail them to rcn-ee so that he can include them in the latest kernel. Jul 08 15:47:15 nerdboy: alexanderhiam on libsysfs, I've got this linked from SO -> http://www.mjmwired.net/kernel/Documentation/sysfs-rules.txt Jul 08 15:47:21 not sure how it could be worse than nothing but... Jul 08 15:47:32 yeah, that's what I was reading neemo Jul 08 15:47:33 rule number one is not to use libsysfs :? Jul 08 15:47:54 rohith: good, can you gist your current status? Jul 08 15:48:23 anujdeshpande: I am planning to first complete all the apps then I will update github wiki and make some tutorial videos Jul 08 15:48:23 vmayoral: Okay. Will also post the errors during compilation. Jul 08 15:48:24 that's kind of, specific, and not a very good first hint Jul 08 15:48:48 http://sourceforge.net/projects/linux-diag/files/ last sysfsutils release was 2006 Jul 08 15:48:49 but that's implementation, although that will come in play soon enough Jul 08 15:49:01 * neemo (gulp) Jul 08 15:49:14 use udev interface i guess Jul 08 15:49:29 or just read/write your sysfs entries directly Jul 08 15:49:36 sounds like libsysfs should just go away... Jul 08 15:49:38 ankur: one usually tends to forget some tiny details over time :) you could just start putting some basic stuff up, and then spend some quality time making it pretty and thorough at the very end. Jul 08 15:50:00 alexanderhiam: that might be a bit more error-prone Jul 08 15:50:14 but that's what they're there for! Jul 08 15:50:21 rohith: sure, we will help you to fix it Jul 08 15:50:23 * nerdboy leery of hacky/homegrown sysfs interface Jul 08 15:50:57 nerdboy: you might want to steer clear of the BeagleBone then! :P Jul 08 15:51:04 since they recommend udev "abstraction" i would look there first Jul 08 15:51:14 also hal is dead long time Jul 08 15:51:29 ahcorde: Thanks. Will post error log and inform you all. Jul 08 15:51:39 alexanderhiam: leery of creating YAHI Jul 08 15:52:20 YAHI? Jul 08 15:52:29 * nerdboy not sure that's a real acronym Jul 08 15:52:36 vmayoral ahcorde: Had to add config, defines and Parameters of Rover to get past the initial few errors. Jul 08 15:52:40 anujdeshpande: ya, I will now make a blog post whenever I will develop any app Jul 08 15:52:41 Yet Another Hacky Interface Jul 08 15:52:49 ah Jul 08 15:52:51 The header files... Jul 08 15:53:07 awesome Jul 08 15:53:25 also, make sure you post it on the beagleboard community on google plus Jul 08 15:53:29 well, everything else is just using sysfs directly Jul 08 15:53:32 well you have this YAHI -> http://rolandturner.com/yahi/, not the same though Jul 08 15:53:55 but, I'm also for the non YAHI variety Jul 08 15:54:08 vmayoral ahcorde: Should I write a blog post for the current progress/status? Haven't updated it in quite some time... Jul 08 15:54:17 I guess I'll look into what all the cool kids are using, and do that Jul 08 15:54:27 good morning Jul 08 15:54:39 * alexanderhiam is missing a quantifier there; everything else I've seen on the BeagleBOne Jul 08 15:54:46 rohith: blog posts are up to you but make sure you update BeagleBoard's official channels Jul 08 15:54:49 mornin moto-timo Jul 08 15:54:57 (e.g.: the mailing list) with weekly reports Jul 08 15:55:08 azizulhakim: good progress Jul 08 15:55:22 vmayoral: Yes. I'm doing that on a weekly basis. Jul 08 15:55:29 thanks vvu :) Jul 08 15:55:29 people too lazy to use udev? Jul 08 15:55:50 rohith: that's enough from my side then Jul 08 15:56:20 Hey rcn-ee Jul 08 15:56:51 Hey Abhishek_ Jul 08 15:57:04 nerdboy: I forget the "rules" on what's a device node Jul 08 15:57:23 rcn-ee: Just got your mail. I'll try pulling from the main repo tonight and testing that hdmi audio. Jul 08 15:57:34 seems like sysfs is safer for more abstracted things that aren't just a single raw device Jul 08 15:57:36 nerdboy: I have no idea Jul 08 15:58:22 hendersa, no rush, i'm out friday too, so i won't be back till sunday.. but with the 4.1.x switch over to ti's tree, we get those extra features.. (it might be just enough for you to move to v4.1.x) Jul 08 15:58:49 maybe spend a few minutes looking at current stuff (sysfs/udev) and see what the code looks like Jul 08 15:59:17 rohith: I'd like to see daily updates Jul 08 15:59:37 pick a decent example and we can review/discuss it Jul 08 16:00:02 your system should have lots of them... Jul 08 16:00:04 vmayoral: Sure. Shall I update in beaglepilot2/meta? Jul 08 16:00:16 I'm looking through libudev stuff Jul 08 16:00:18 <_av500_> aloha Jul 08 16:00:22 <_av500_> 5min to launch Jul 08 16:00:58 everybody ready to yank the "Pull before flight" tags? Jul 08 16:01:23 rohith: I'm interested more in the contributions to the vehicle. Reviewed your commits and i think that's not enough to meet the schedule. Jul 08 16:01:34 I went through this guide http://derekmolloy.ie/kernel-gpio-programming-buttons-and-leds/, unfortunately he doesn't tackle the how to interface with sysfs part Jul 08 16:01:50 vvu: have you been thinking about the flash system now that university is over? Jul 08 16:01:55 of to more oggling, be back for the launch Jul 08 16:02:19 vmayoral: I tried to get the code compiling. Have been working on the steering mechanism too. Jul 08 16:02:28 hi all Jul 08 16:02:45 good morning jkridner Jul 08 16:02:49 hola jkridner Jul 08 16:02:53 vmayoral: Will focus more on the submarine part. Jul 08 16:02:59 hi jkridner Jul 08 16:03:03 Hey jkridner Jul 08 16:03:12 greetings jkridner Jul 08 16:03:32 _av500_: who is ‘presenting’ today ? Jul 08 16:03:49 ankur: do you have a slot yet ? Jul 08 16:04:21 moto-timo: not so much Jul 08 16:04:31 what do we need exactly now? Jul 08 16:04:48 anujdeshpande: no not yet. may be next week Jul 08 16:04:50 Hello all Jul 08 16:04:53 linux and osx we are set...windows i have not started Jul 08 16:05:01 <_av500_> anujdeshpande: rohith and arianepaola_ Jul 08 16:05:05 <_av500_> are up today Jul 08 16:05:05 here Jul 08 16:05:06 I believe arianepaola_ should be presenting Jul 08 16:05:13 <_av500_> yes Jul 08 16:05:16 <_av500_> time is up Jul 08 16:05:19 yes jkridner Jul 08 16:05:23 <_av500_> 1) WELCOME everybody Jul 08 16:05:24 * moto-timo goes silent Jul 08 16:05:30 Attendance Jul 08 16:05:36 <_av500_> 2) who's here? Jul 08 16:05:40 here Jul 08 16:05:42 here Jul 08 16:05:43 here Jul 08 16:05:46 here Jul 08 16:05:53 here Jul 08 16:06:10 here Jul 08 16:06:15 here Jul 08 16:06:26 <_av500_> 7 Jul 08 16:07:11 <_av500_> ah Jul 08 16:07:12 <_av500_> #8 Jul 08 16:07:21 <_av500_> hi ebadawy Jul 08 16:07:37 hi _av500_, sorry for being late :) Jul 08 16:07:41 <_av500_> np Jul 08 16:07:47 <_av500_> 3) congrats for passing midterms Jul 08 16:08:09 thanks :) Jul 08 16:08:28 o7 Jul 08 16:08:41 received the midterm payment as well :) Jul 08 16:09:35 <_av500_> like midterms, finals are there fester than you think Jul 08 16:09:39 <_av500_> faster* Jul 08 16:09:42 <_av500_> so dont lose steam Jul 08 16:10:16 <_av500_> 4) blocking issues? Jul 08 16:10:47 not yet Jul 08 16:11:04 _av500_ have one Jul 08 16:11:10 nope Jul 08 16:11:13 still stuck on that equation solver, working on other parts in the meantime Jul 08 16:11:15 no issue Jul 08 16:11:17 no issue Jul 08 16:11:44 nope Jul 08 16:11:52 no Jul 08 16:12:06 <_av500_> ebadawy: DiegoTc and jkridner are aware? Jul 08 16:12:42 jkridner: DiegoTc told me yesterday that BBUI will be considered as another type of bonecard, is it the case? Jul 08 16:12:48 _av500_: not aware of one that is blocking.... ebadawy: please elaborate. Jul 08 16:12:53 ebadawy: correct. Jul 08 16:13:12 ok, that's it :) Jul 08 16:14:38 <_av500_> good :) Jul 08 16:15:12 <_av500_> 5) status updates Jul 08 16:15:24 <_av500_> arianepaola: you have the stage Jul 08 16:15:35 _av500_: thank you Jul 08 16:16:02 my project is to create a beaglebone getting started application Jul 08 16:16:18 that is crossplatform and includes several features that should make life easier for new users Jul 08 16:16:54 the idea is to allow inexperienced users to start using beablebone quickly and without issues Jul 08 16:17:21 so that they do not have to install a ssh client, configure network settings, install programs for flashing or even have to download the flash images on their own Jul 08 16:18:03 This project is based around the getting started guide that is currently available and automates the features that are displayed there, like connecting the beaglebone, flashing, etc. Jul 08 16:18:16 this is explained in more detail in the video presentation. Jul 08 16:18:49 do you have any questions so far? or can I continue to talk about the technologies used? Jul 08 16:19:16 what do you see as the main reason someone would download this vs. using a web browser? Jul 08 16:19:35 just as a note here is the presentation: https://www.youtube.com/watch?v=XVZS8kLhhwY Jul 08 16:19:43 ebadawy: told you Jul 08 16:20:01 DiegoTc: yeah :D Jul 08 16:20:01 quite know jkridner thoughts on his view of bone101 Jul 08 16:20:13 jkridner: it would be possible to ship the application on the beagle sd card and to make it available to the user, including automation of tasks such as driver installation Jul 08 16:20:20 it is there to make things easier Jul 08 16:20:41 when someone is new to the beaglebone it might be difficult to find the correct tools needed, setup the configuration, etc. Jul 08 16:20:55 or even if you are lazy, the application is like google for beagle :-) Jul 08 16:21:03 but everybody knows how to open a thumb drive :) Jul 08 16:21:04 it provides a starting point Jul 08 16:21:06 :) Jul 08 16:21:18 DiegoTc: great, coz I may need to know more things about BBUI, but let me start with it first Jul 08 16:21:38 moto-timo: I did not understand this about the thumb drive Jul 08 16:22:02 like a usb flash drive... Jul 08 16:22:16 of course the application has some nice features, such as internationalization which is included and if you do not speak english, you will be able to change the language to understand better what you are doing Jul 08 16:22:42 moto-timo: do you mean that the drivers are on the beagle flash? Jul 08 16:23:20 moto-timo: still if the user gets to install those, what about flashing and connecting? if you are a newbie, the fun might be over very quickly Jul 08 16:23:43 when you connect the BBB with usb cable, a FATFS is mounted... if the application is there it is easy to launch Jul 08 16:23:57 arianepaola: is the flashing step really coming together with vvu? do you feel this app will make it usable? Jul 08 16:24:08 moto-timo: yes this is what I meant, it would be great to include it on the flash Jul 08 16:24:17 :) Jul 08 16:24:22 jkridner: this is the next thing that I am going to work on this week and in the next weeks Jul 08 16:24:34 * moto-timo worked on it early to get ready Jul 08 16:25:49 may I go on? Jul 08 16:25:54 arianepaola: please. Jul 08 16:26:18 so to the technical part Jul 08 16:26:44 probably everyone knows nodejs or has heard of it. it is an event driven server side javascript environment Jul 08 16:27:19 it has become very popular recently and for this project, I am relying on nwjs (before called node webkit) Jul 08 16:27:50 nw.js allows you to use the engine of chromium (open source part of google chrome) to render the application by using standard html, css, javascript, etc. Jul 08 16:28:09 and the best part is that you can directly access nodejs modules Jul 08 16:28:25 although they are severside modules you can just run them on top of nwjs with client side code Jul 08 16:28:43 nwjs allows you to create portable applications for linux, mac and windows Jul 08 16:29:03 any questions so far? Jul 08 16:29:24 you publishing screenshots as you create them? Jul 08 16:29:38 how are you thinking about usability. Jul 08 16:29:56 yes, there are some screenshots in the video presentation Jul 08 16:30:18 for usability, I have thought of making things easy to find and offering a limited set of options to the user Jul 08 16:30:30 big buttons with big text, so things do not go wrong Jul 08 16:30:42 and menu entries like: Connect Jul 08 16:30:51 that make it easy to understand what the users are looking for Jul 08 16:31:39 for usability, design also plays an important part, I am using bootstrap (before twitter bootstrap), a frontend framework, that makes things look nice. Jul 08 16:31:54 arianepaola: are you using any content from bone101 or just beaglebone-getting-started? Jul 08 16:31:59 they have very clean and organized user interface elements Jul 08 16:32:06 I don't think I've seen any pull requests on it lately. Jul 08 16:32:27 did you and moto-timo ever work out an automated build process where people can download your nightly builds? I still haven't seen a binary to try. Jul 08 16:33:08 jkridner: I have looked at parts of bone101 to better understand some parts of socket.io and how communication works Jul 08 16:33:15 I did not reuse any code specifically Jul 08 16:33:39 jkridner: can you recommend any specific parts to look at? Jul 08 16:34:08 jkridner: I did update the config for travis ci and also set up appvoyer (still gives me errors) Jul 08 16:34:28 at the moment I am still using grunt build, but it does the complete process Jul 08 16:34:48 arianepaola: the tutorials interact with the board, so looking at those should be something you do. I think 'uploader' only exists on beagleboard.org right now, but it is a hack. Jul 08 16:35:41 jkridner: thanks Jul 08 16:35:55 so as I have mentioned the grunt builder Jul 08 16:36:10 with the latest commits to github, I have something to work with as far as appveyor and travis ci builds Jul 08 16:36:26 thanks moto-timo Jul 08 16:36:57 so grunt is a javascript task runner. you can think of it like a very nice and modern make and automation system Jul 08 16:38:10 I have added several tasks to the builder to create the application, download dependencies, run the translations, run tests, build the packages, among others Jul 08 16:38:31 <_av500_> 2min Jul 08 16:38:34 <_av500_> :) Jul 08 16:38:40 ok, really quick then Jul 08 16:38:50 so the nw.js part is based on html / css for the frontend Jul 08 16:39:07 a difficulty of course is the translation, therefore jade (a template language for html) is being used Jul 08 16:39:26 there is one template for the frontend written in jade, an a set of translation files Jul 08 16:39:44 the build scripts converts for each build everything to make it nice and shiny and translated Jul 08 16:39:52 then when we go into the features Jul 08 16:40:16 the detection of the beaglebone is done or statically via the default ip or using mdns to autodetect the board Jul 08 16:40:23 this information is used to connect to the board Jul 08 16:41:07 arianepaola: do you get them connected to cloud9? Jul 08 16:41:10 it is possible to connect to the board from the application, using a terminal application, which talks to a socket.io server internally, sending all commands, this server does the real ssh connection behind the scene, talking to the beaglebone and executing the commands Jul 08 16:41:27 I am using ssh2 and scp2 at the moment Jul 08 16:41:40 arianepaola: why not leverage the cloud9 app already on the board? Jul 08 16:41:58 jkridner: does it come by default? Jul 08 16:42:03 arianepaola: yes Jul 08 16:42:07 jkridner: lol Jul 08 16:42:21 jkridner: I was reinventing the wheel Jul 08 16:42:52 :( Jul 08 16:42:55 jkridner: I can add an option to the menus to connect to the cloud9 app on the board Jul 08 16:42:56 <_av500_> arianepaola: time Jul 08 16:42:57 sorry to not catch it sooner. Jul 08 16:43:08 maybe her wheel is better... Jul 08 16:43:11 so there is also a scp feature, at the moment only drag and drop Jul 08 16:43:19 <_av500_> arianepaola: we have to wrap it up now Jul 08 16:43:21 so you drag a file on to the app window and it copies it to the board Jul 08 16:43:23 ok Jul 08 16:43:27 <_av500_> thanks Jul 08 16:43:31 <_av500_> rohith: you turn Jul 08 16:43:48 _av500_: Sure Jul 08 16:44:02 <_av500_> ankur: ebadawy: you ok for next week? Jul 08 16:44:12 <_av500_> rohith: go on Jul 08 16:44:12 yes Jul 08 16:44:16 yes Jul 08 16:44:18 My project is BeaglePilot 2.0 - Making underwater drones Jul 08 16:44:32 Code: https://github.com/beaglepilot2 Jul 08 16:44:57 The aim of the project is to create an open hardware underwater vehicle Jul 08 16:45:05 using the BeagleBone Black Jul 08 16:45:21 It involves building on the ardupilot code Jul 08 16:45:40 https://github.com/diydrones/ardupilot Jul 08 16:45:57 to use its existing autopilot capabilities Jul 08 16:46:17 ie, to extend APM to include a new vehilce Jul 08 16:46:24 the APMsubmarine Jul 08 16:46:52 The project will enable hobbyists to get access to a low cost, easy to use Jul 08 16:47:13 and open hardware underwater vehicle platform Jul 08 16:47:22 with autopilot capabilities Jul 08 16:47:37 Any questions so far??? Jul 08 16:48:13 <_av500_> go on :) Jul 08 16:48:21 Okay. Jul 08 16:48:49 OpenROV is currently the only open source underwater vehicle system Jul 08 16:49:06 but has a slightly complex structure Jul 08 16:49:22 considering that commands are sent via nodejs Jul 08 16:49:41 to a separate Arduino based controller boardcape Jul 08 16:49:59 Also, it is fairly basic in operation Jul 08 16:50:23 and has basic teleoperation + video feed system Jul 08 16:50:53 It does not have features such as speed estimation Jul 08 16:51:08 and also no autopilot mode, Jul 08 16:51:13 just a manual one. Jul 08 16:51:29 We would like to add more capabilities Jul 08 16:51:44 such as the aforementioned autopilot Jul 08 16:51:54 and even speed estimation later Jul 08 16:52:10 <_av500_> can you tell us about your progress? Jul 08 16:52:14 <_av500_> and achievements? Jul 08 16:52:23 but the idea is to first achieve a working/controllable vehicle Jul 08 16:52:32 _av500_: Okay Jul 08 16:52:43 What has been done so far - Jul 08 16:53:34 I have researched and understood the functioning of the OpenROV system and ardupilot to an extent Jul 08 16:53:57 and also added several files to the APMsubmarine vehicle Jul 08 16:54:04 https://github.com/beaglepilot2/ardupilot Jul 08 16:54:19 and also a motor library for the same. Jul 08 16:54:38 I've used APMrover2 as a reference Jul 08 16:54:49 http://rover.ardupilot.com Jul 08 16:55:07 it being closest to a submarine in operation Jul 08 16:55:27 since the rear wheel motors of the rover Jul 08 16:55:52 can be correlated to the port and starboard thrusters on the submarine/ROV Jul 08 16:56:31 One issue faced during this process was that the OpenROV currently does not have a dynamic model implemented Jul 08 16:57:02 which would have been required to be compatible fullt with the existing APM vehicles Jul 08 16:57:21 rohith: what code did you start with? Jul 08 16:57:28 rohith: I mean, what did you write first? Jul 08 16:57:32 as they incorporate the dynamic model in the code for location and navigation Jul 08 16:58:08 jkridner: I started with the basic files that constitute any APM vehicle Jul 08 16:58:42 ie, the cpp files for the vehicle Jul 08 16:59:13 The approach now is to implement the steering mechanism Jul 08 16:59:41 which I've understood and mentioned in my Notes.txt file Jul 08 17:00:08 https://github.com/BeaglePilot2/ardupilot/blob/master/Notes.txt Jul 08 17:01:12 rohith: did you add any functionality to it? get the fancy sensors working? Jul 08 17:02:36 * ebadawy have to go now for the breakfast after a fasting day! Jul 08 17:04:26 <_av500_> rohith: go on Jul 08 17:04:55 <_av500_> I guess we lost him Jul 08 17:05:01 <_av500_> also, we're late Jul 08 17:05:12 <_av500_> so, time to wrap it up Jul 08 17:05:16 <_av500_> thanks all Jul 08 17:05:21 <_av500_> see you next week Jul 08 17:05:23 <_av500_> happy coding Jul 08 17:05:45 thanks _av500_ Jul 08 17:06:12 I got disconnected. Sorry. Jul 08 17:06:31 What was the last thing you received? Jul 08 17:06:51 00:52 < rohith> https://github.com/BeaglePilot2/ardupilot/blob/master/Notes.txt Jul 08 17:07:04 neemo: Thanks :) Jul 08 17:07:29 Also, I've tried, albeit unsuccessfully, to compile the vehicle code Jul 08 17:07:38 something I hope to solve in the coming week Jul 08 17:07:46 You would also notice that part of the code is rover based Jul 08 17:07:53 It is because a parallel could not be found yet for the underwater vehicle Jul 08 17:08:00 and was added to test if the vehicle code compiles Jul 08 17:08:06 Any questions? Jul 08 17:08:31 * rohith had been typing but had got disconnected :( Jul 08 17:10:55 * jkridner is about to drop. Jul 08 17:16:14 Apologies again to all. Seems like the meeting is over :( Jul 08 17:17:36 rohith: http://logs.nslu2-linux.org/livelogs/beagle-gsoc.txt Jul 08 17:18:10 Thanks ankur_ Jul 08 17:18:23 Went through it... Jul 08 17:27:21 nerdboy: alexanderhiam I'm gonna go sheep counting soon Jul 08 17:28:04 nerdboy: alexanderhiam meet up here tomorrow at around (this time - 1hr) Jul 08 17:28:20 e.i. today's meeting time Jul 08 17:28:21 ? Jul 08 17:29:27 or sooner... Jul 08 17:29:40 or sooner Jul 08 17:30:12 but I'll have to sink some time at the visa office tomorrow morning Jul 08 17:30:16 you don't get to look at my solver until you get through some of this stuff... Jul 08 17:30:27 and sleep is in between now and then Jul 08 17:30:44 do what ya gotta do Jul 08 17:30:54 nerdboy: hah, did you solve those equations? Jul 08 17:31:13 that's your job Jul 08 17:31:14 I'll be working on the API, no detou Jul 08 17:31:16 detours Jul 08 17:31:37