**** BEGIN LOGGING AT Wed May 07 02:59:58 2014
May 07 04:29:56 <sidbh> hi adamgamble. pls checkout this mailling list https://groups.google.com/forum/#!forum/beaglepilot, it will tell you everything you need to know & contribute
May 07 07:55:50 <sid_> panto: ping
May 07 12:35:00 <Not-6d9> [beaglepilot] vmayoral pushed 1 commit to master [+0/-0/±1] http://git.io/x3GyVw
May 07 12:35:01 <Not-6d9> [beaglepilot] vmayoral b78864b - Weekly report update. New sensors added, previous drivers work after some fixes. Barometer doesn't seem to work properly.
May 07 13:07:51 <adamgamble> thanks sidbh
May 07 13:51:13 <vmayoral> adamgamble: just saw your message, glad you are joining our efforts!
May 07 14:00:49 <adamgamble> Yea i've been wanting to contribute to open source stuff for awhile
May 07 14:01:01 <adamgamble> so far I only have commits in Ruby on Rails
May 07 14:01:25 <vmayoral> just got your message on the grou
May 07 14:01:26 <vmayoral> p
May 07 14:01:48 <vmayoral> the web interface you refer is (as planned) a Web-based IDE
May 07 14:01:56 <adamgamble> oh ok gotcha
May 07 14:02:06 <vmayoral> so far Cloud9 is working fine in the BeagleBone but I also plan to test Codiad
May 07 14:02:24 <vmayoral> there's quite a bit of work to do an that side though
May 07 14:02:24 <adamgamble> Are there any plans for a web based apm planner type project?
May 07 14:03:32 <vmayoral> there's already something like that, check Mavelous (http://github.com/wiseman/mavelous)
May 07 14:03:44 <vmayoral> we were planning to start from there actually
May 07 14:03:57 <adamgamble> excellent
May 07 14:04:03 <adamgamble> yep thats exactly what I meant
May 07 14:04:07 <vmayoral> would you like to contribute that way?
May 07 14:04:13 <vmayoral> that'll be great actuallyu
May 07 14:04:20 <adamgamble> yea I would love to
May 07 14:04:36 <adamgamble> I'll have to brush up on my python, but that shouldn't be an issue
May 07 14:05:00 <vmayoral> :), If you check our initial proposal (http://github.com/beaglepilot/beaglepilot), you'll see that Anuj will be working at the final weeks on GCS
May 07 14:05:16 <vmayoral> he's not around now, but i'm having a meeting with him later on so i can come back to you
May 07 14:05:26 <adamgamble> ok excelletn
May 07 14:05:35 <adamgamble> in the meantime i'll get that project running locally
May 07 14:05:53 <vmayoral> sure, that'll be a great starting point. I'm not sure in which GCS he will focus
May 07 14:06:08 <vmayoral> if i had to choose i would go for droidplanner or Mavelous
May 07 14:06:19 <vmayoral> both are looking great, probably the first one has a broader support
May 07 14:06:55 <vmayoral> so your contribution will be great :)!
May 07 14:07:02 <adamgamble> so is the plan to not fall in line with whatever supports ardupilot will also support beagle pilot?
May 07 14:07:17 <adamgamble> I figured all the mavlink stuff would be the same
May 07 14:07:32 <vmayoral> yes
May 07 14:07:34 <adamgamble> or am I not understanding what you mean by picking a specfic GCS
May 07 14:07:53 <adamgamble> just planning on having other new commands that wont be supported by ardupilot?
May 07 14:08:02 <adamgamble> or just an official supported route
May 07 14:08:29 <vmayoral> the last one, an official, supported, tested and documented way
May 07 14:08:37 <adamgamble> gotcha
May 07 14:08:41 <adamgamble> we're on the same page now
May 07 14:08:46 <vmayoral> :)
May 07 14:09:08 <adamgamble> well awesome, i'm pumped to help out, I've been having a blast with my quad
May 07 14:09:32 <vmayoral> they are fun, indeed :). We are really excited about the idea of having linux machines flying out there
May 07 14:11:17 <vmayoral> Settled then, 'll discuss the Mavelous matter with Anuj and come back to you later tonight.
May 07 14:11:18 <adamgamble> yea and I think it will awesome and be way more extensible, then you could script it ever how you wanted or run arbitrary processes
May 07 14:13:17 <adamgamble> ok sounds great
May 07 14:22:36 <adamgamble> Did you guys get selected for GSOC?
May 07 15:46:01 <vmayoral> adamgamble: i did get selected
May 07 15:46:46 <adamgamble> so i've been looking over the code for mavelous
May 07 15:48:48 <adamgamble> since that codebase is currently unmaintained I was thinking about starting from scratch (even if it isn't the direction you guys decide to go on this project)
May 07 15:49:02 <adamgamble> I think I could get to feature parity with mavelous fairly quickly
May 07 15:49:11 <adamgamble> ~ a weekend
May 07 15:49:58 <vmayoral> sounds good on me
May 07 15:50:14 <vmayoral> i also saw that the project has been stucked for the las 7 months
May 07 15:50:16 <adamgamble> so if yall are interested I would be open to just starting a project that I maintain and is our official supported GCS, but I'm not trying to call the shots, I'll go whatever route you guys want to go
May 07 15:50:20 <adamgamble> yep
May 07 15:54:35 <vmayoral> like the idea and i don't think that Anuj has advanced much on the GCS side because now he's focused on the the DT which is the main priority.
May 07 15:55:26 <adamgamble> well with me helping it might could free him up to other parts of the project
May 07 15:55:35 <adamgamble> I know the schedule is tight
May 07 15:55:46 <vmayoral> yeap
May 07 15:56:31 <vmayoral> i'm personally also interested in GCS so i can give a hand now and then
May 07 15:57:00 <adamgamble> do you have a language prefence?
May 07 15:57:03 <adamgamble> preference
May 07 15:57:10 <adamgamble> I'm a ruby guy
May 07 15:57:20 <adamgamble> but can be whatever
May 07 15:58:10 <adamgamble> was thinking a light webserver to serve the web app, then a web socket to stream the data back and forth
May 07 15:58:18 <vmayoral> i usually prefer python but i've also worked with ruby, no problem on my side
May 07 15:58:33 <adamgamble> are you full stack or backend?
May 07 16:00:13 <adamgamble> Basically I was thinking of building a javascript app on the front end that consummed the websockets, populated its models and those models were bound to the view
May 07 16:00:54 <adamgamble> again though I don't want to get ahead of ourselves, if you guys are wanting me to try to maintain some other project
May 07 16:00:56 <vmayoral> i like to think of myself as full stack but i can't compromise myself (although i will try to push as well) because i have to accomplish the GSOC goals that i created (already a bit challenging)
May 07 16:01:21 <adamgamble> completely understand
May 07 16:01:32 <vmayoral> i was wondering, do you believe that the server side could work on the bone itself?
May 07 16:02:02 <adamgamble> I do, though I don't actually have a BBB yet (it will be here friday I think)
May 07 16:02:22 <adamgamble> I have a pcduino though and i've run several apps on it that published data over websockets
May 07 16:02:42 <adamgamble> it will depend how much CPU/RAM the rest of the code is taking up though
May 07 16:03:01 <adamgamble> I imagine the beaglepilot code will be fairly taxing unless you guys are offloading most to the fire cape
May 07 16:03:23 <vmayoral> yes, that's what i'm afraid as well
May 07 16:03:34 <adamgamble> so making it light weight is definitely a priority
May 07 16:04:06 <adamgamble> how close are you guys to having one flying?
May 07 16:04:23 <vmayoral> yeap, i will try to provide data about the computational needs of beaglepilot asap but i keep programming drivers for now
May 07 16:04:50 <vmayoral> mmm well drivers for a minimum flight should be coded but there are still some issues to address (and many that keep appearing)
May 07 16:05:13 <vmayoral> i have with myself a BB-based robot, Erle (http://erlerobot.com) that is ready for flight the moment beaglepilot is
May 07 16:05:24 <adamgamble> honestly we could do most of the processing on the front end and leave the backend to really just pipe the mavlink stuff up the websocket
May 07 16:05:29 <adamgamble> that would be very lightweight
May 07 16:05:48 <adamgamble> and it isn't like there is going to be more than 1 or so websockets connected, it doesn't need to scale
May 07 16:06:52 <vmayoral> the PRU code for outputing PWM should take a while as well so eventually the eHRPWM modules can be used.... difficult to estimate how much time but i hope that soon :)
May 07 16:07:17 <vmayoral> regarding offloading the computation and putting it all in the front end. Perfect!
May 07 16:07:29 <panto> vmayoral, err, there's code for that already
May 07 16:08:09 <vmayoral> panto: hi! yeap sure but we might need to wire it with ardupilot
May 07 16:08:48 <vmayoral> panto: not saying that that's hard just that it might take time, the code is reasonably complex (about half a million lines)
May 07 16:09:00 <adamgamble> so basically i'm thinking a lightweight process on the backend that can both deliver mavlink data to whatever processes you guys have and send it up to the front end
May 07 16:09:02 <panto> no no no :)
May 07 16:09:08 <panto> I have to answer to that damn email
May 07 16:09:18 <panto> just so busy this couple of days
May 07 16:09:54 <vmayoral> adamgable: is it possible to design everything with multi-uavs in mind?
May 07 16:10:10 <vmayoral> please take this into account if you are putting the computation in the front end
May 07 16:10:25 <adamgamble> hmm
May 07 16:11:03 <adamgamble> well I wasn't thinking of that no but I guess in theory we could build it so it could connect to multiple websockets on different BBB
May 07 16:11:16 <adamgamble> and the process would roughly be the same
May 07 16:12:02 <adamgamble> I would need a way to plug in to whatever processes you guys have running and grab the telemtry on the BBB
May 07 16:12:42 <adamgamble> how is that currently setup? Are you just writing directly to the serial port for all the telemtry or is there some sort of daemon that routes it to multiple places if need be
May 07 16:19:43 <vmayoral> there's a serial port but i'd say you can just do it through ROS
May 07 16:19:54 <vmayoral> i'm supposed to integrate ROS within ardupilot in a few weeks
May 07 16:20:59 <adamgamble> ok
May 07 16:21:26 <adamgamble> so basically the backend would be a light server to serve the files + a websocket of mavlink data
May 07 16:22:37 <adamgamble> so keep in mind none of this makes sense unless its in the context of being on a network with the BBB
May 07 16:23:00 <adamgamble> so it wouldn't work with the 3dr radios
May 07 16:23:26 <vmayoral> not familiar with the 3DR radios but i see the logic behind our reasoning
May 07 16:24:02 <adamgamble> would it be possible to setup a google hangout or skype call or something sometime to discuss all this?
May 07 16:24:05 <vmayoral> regarding how to get the telemetry there're different ways, either through the serial port or reading the ROS topics
May 07 16:24:19 <vmayoral> sure
May 07 16:24:22 <adamgamble> I don't want to pull us in a direction that doesn't meet the goals you guys have that i'm possibly not thinking of
May 07 16:24:46 <adamgamble> I'm pretty much good any time except about 6 - 8:30 central tonight
May 07 16:24:47 <vmayoral> how about you state these ideas in a new thread and we let the others jump in and share ideas?
May 07 16:24:54 <adamgamble> ok that sounds good.
May 07 16:26:45 <vmayoral> i'm also fine with a hangout or a skype call but i'd suggest we wait a few days so that the people can come up with ideas. We are growing fast and many people might have interesting pointers
May 07 16:27:03 <adamgamble> yea I agree that sounds like the right call
May 07 16:27:36 <adamgamble> I'll come up with a post and post it to the group, I want to get everyones input because I don't want to just jump in and start coding and realize I went in the wrong direction
May 07 16:27:47 <adamgamble> or that I didn't support some needed use case
May 07 16:30:48 <vmayoral> yeap, that sounds just the right way to go :). Happy you joined Adam!
May 07 16:31:00 <adamgamble> thanks me too!
**** ENDING LOGGING AT Thu May 08 02:59:58 2014