I think our best opportunity to recruit good mentorship for Google Summer of Code 2019 is to define some projects that will really benefit the community and solve challenges those individual mentors would love to see solved so they can be more productive in the future. Mentors also just love playing with new tech. I’d really appreciate your feedback on what those projects should be.
I’ve added my first idea to the ideas page!
https://elinux.org/BeagleBoard/GSoC/Ideas#click_board_support_under_Greybus
click board support under Greybus
Today, support for click boards on PocketBeagle is accomplished via device tree overlays via the bb.org-overlays repository. This requires /boot/uEnv.txt to be modified to load the drivers at boot, requiring at least one reboot to enable the support in a potentially error-prone way, as modifying that file could prevent the image from booting. Fortunately, some work on BeagleBone to produce a Greybus simulator opens up a different possibility. Greybus already provides most of the interfaces used on click boards and utilizes manifest files to enumerate hardware at run-time. By using the simulator, similar functionality should be possible to achieve for these existing buses, rather than via true hot-swapping hardware.
Goal: Greybus drivers, userspace simulator and example manifests for several MikroElectronkia click boards using PocketBeagle.
Hardware Skills: Some preliminary understanding of some basic interface protocols like UART, I2C, and SPI
Software Skills: Linux kernel modules
Possible Mentors: Jason Kridner
Upstream Repository: https://github.com/beagleboard/linux, https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc
References: https://lwn.net/Articles/715955/
Learn more at https://beagleboard.org/gsoc
What is Google Summer of Code?
“Google Summer of Code is a global program that offers students stipends to write code for open source projects. We have worked with the open source community to identify and fund exciting projects for the upcoming summer.” - **
Status
We are planning to apply as a mentoring organization for GSoC 2019. You encouraged to join us on http://bbb.io/gsocchat if you want to get involved. Administrators are ‘av500’, ‘jkridner’, and ‘cwicks’.
Step-by-step:1. Sign up on the mailing list at http://bbb.io/gsocml
- Join the live chat at http://bbb.io/gsocchat (#beagle-gsoc) including weekly meetings every Wednesday at noon US Eastern time (http://bbb.io/gsocmeet), which is 4PM UTC during US daylight savings time.
- Read and contribute to our ideas list at http://bbb.io/gsocideas.
- Sign up on Google’s system at https://summerofcode.withgoogle.com/
can us students/participants add ideas too at this stage or is this only limited to mentors. I have cool(i think so) ideas that mentors can pickup ?
can us students/participants add ideas too at this stage or is this only limited to mentors. I have cool(i think so) ideas that mentors can pickup ?
It is fine. If it becomes a problem (too many to review), then we’ll change the policy.
Letting us know a bit about the idea on this thread is also good.
can us students/participants add ideas too at this stage or is this only limited to mentors. I have cool(i think so) ideas that mentors can pickup ?
It is fine. If it becomes a problem (too many to review), then we’ll change the policy.
Letting us know a bit about the idea on this thread is also good.
I got 2 ideas ,
1st - Aquaponics using OpenCv,ML ,these systems costs around 400$+(i had one) for basic functionality,with a pretrained image recognition model this could be easily implemented,other than the basic stuff like monitoring temps,ph of the water,running pumps etc it can recognise fish health,ready for harvest size(commercial purposes) and that kind of stuff,isolation for medication.
2nd - PocketBeagle Based StepWise Calculator for Engineering Math, the only thing i am concerned about is the legality of this.Any texas instruments or Casio calc can fit a pocket beagle,with a thin battery in it.It could use a api to send data to wolfgram or sympholab and transcode the solutions step wise on the screen or use mathematical python modules as well.
can us students/participants add ideas too at this stage or is this only limited to mentors. I have cool(i think so) ideas that mentors can pickup ?
It is fine. If it becomes a problem (too many to review), then we’ll change the policy.
Letting us know a bit about the idea on this thread is also good.
I got 2 ideas ,
1st - Aquaponics using OpenCv,ML ,these systems costs around 400$+(i had one) for basic functionality,with a pretrained image recognition model this could be easily implemented,other than the basic stuff like monitoring temps,ph of the water,running pumps etc it can recognise fish health,ready for harvest size(commercial purposes) and that kind of stuff,isolation for medication.
There are some existing efforts I’d want to see how we could support. Namely, https://www.media.mit.edu/groups/open-agriculture-openag/overview/.
2nd - PocketBeagle Based StepWise Calculator for Engineering Math, the only thing i am concerned about is the legality of this.Any texas instruments or Casio calc can fit a pocket beagle,with a thin battery in it.It could use a api to send data to wolfgram or sympholab and transcode the solutions step wise on the screen or use mathematical python modules as well.
Sounds complicated to make and hard to use.
can us students/participants add ideas too at this stage or is this only limited to mentors. I have cool(i think so) ideas that mentors can pickup ?
It is fine. If it becomes a problem (too many to review), then we’ll change the policy.
Letting us know a bit about the idea on this thread is also good.
I got 2 ideas ,
1st - Aquaponics using OpenCv,ML ,these systems costs around 400$+(i had one) for basic functionality,with a pretrained image recognition model this could be easily implemented,other than the basic stuff like monitoring temps,ph of the water,running pumps etc it can recognise fish health,ready for harvest size(commercial purposes) and that kind of stuff,isolation for medication.
There are some existing efforts I’d want to see how we could support. Namely, https://www.media.mit.edu/groups/open-agriculture-openag/overview/.
The open agriculture project/effort looks interesting,soil acidity and grow recommendation computer is something i have thought of before,i will float some ideas here i have written up.
RIght now it seems these openag efforts are targeted towards a closed growing environment, projects in the direction of more own growing would be nice.
Correct me if i am wrong but wasn’t there a open ag computer project last year that was aimed at food storage ? can you share why was it not picked up or what did beagleboard expect.
2nd - PocketBeagle Based StepWise Calculator for Engineering Math, the only thing i am concerned about is the legality of this.Any texas instruments or Casio calc can fit a pocket beagle,with a thin battery in it.It could use a api to send data to wolfgram or sympholab and transcode the solutions step wise on the screen or use mathematical python modules as well.
Sounds complicated to make and hard to use.
Yes it is not the easiest thing to build in terms of the hardware interfacing,especially the keyboard and screen if the usage has to be easy.Maybe this is not suitable for GSOC as It is more of a software thing.
I would love to work on a project that enables the PocketBeagle to be used as an USB keystroke injector. When a host computer is connected, the PocketBeagle registers itself as a USB keyboard+mouse so that it may be able to inject pre-programmed key presses (and mouse clicks). Use cases may range from changing the desktop wallpaper to more malicious tasks (like installing a backdoor on the host!).
(The idea is inspired by the USB rubber ducky)
Can you and all others interested come onto #beagle-gsoc on irc.freenode.net to discuss as well?
The BeagleBoard.org - gsocideas page is a wiki, so it is fine to make reasonable edits there to share and formalize ideas.
I would like to propose the project for GSoC 2019 as a student.
To improve Xen support on Beagleboard-x15. Right now, there is a lot of field for improvement, for example, some time ago there were complaints that Xen didn’t work from the box on top of 4.9 Linux kernel: https://lists.xen.org/archives/html/xen-users/2018-01/msg00119.html. At the same time, Beagleboard hasn’t been even listed in Xen for Arm page: https://lists.xen.org/archives/html/xen-users/2018-01/msg00119.html
So, the motivation behind the project is to make virtualization and experiments with it more available for end users.