Summary/outcome of today’s meeting and way ahead:
mentor concerns in general revolve mainly around scope/effort and the assumptions buried in some of the high-level tasks
in general, everyone needs to start updating task scope and try and make it more detailed
in some cases (for example) there probably needs to be some time for algorithm/library testing and a backup plan if plan A blows up completely- should not assume writing something non-trivial from scratch is the best plan when there are already tested/proven implementations
if/when it starts to look bigger than you thought you’ll want to look at trimming down your baseline and maybe moving some tasks to “stretch” goals (which you can work on if time permits, or put off as post-GSoC tasks
So if your current project timeline/schedule has stuff like “Week 3 - implement x” then please break that task into smaller chunks. If your plan contains Big Untested Assumptions about performance, library interfaces, code size for the PRUs, etc, then you better think about adding some time to evaluate/analyze/test your assumptions (and have a backup/alternate plan ready). Make sure you have some time to write and execute tests on anything you write/modify (yes there is guidance for that).
Please update your timelines as needed.
Also, many of you are planning to work with specific kernel interfaces/subsystems so make sure you understand the two main kernels available and their -rt (real-time) variants. You should be looking at and/or building against 4.4 bb-kernel or 4.1 linux-ti-dev (both maintained by Robert Nelson and packaged in the stock beaglebonre Debian and Ubuntu images). The capemgr README gives a brief overview of this, and Robert’s wiki page has the manual build/deploy process and configuration tips; see the capemgr: v4.1.x+ section for example.
Start using the package repos in your beaglebone and choose one of these kernels. Ask questions if things are too confusing…
Steve aka nerdboy