Tuesday, 16 July 2013

GSOC Update 2

This blog post covers the progress for my RTEMS GSOC Project Unified APIs from 5th July till now( roughly the next 10 days, since my last GSOC post ). The task for removing legacy support completely from different BSPs is on, and in full swing.

As of now, major work has been done on ARM BSPs which include csb336, csb337, edb7312, gumstix, rtl22xx & gba. I've been asked to set the nds BSP aside for a while, since it uses custom internal methods defined for that BSP, but having the same name as the Legacy API. I'm right now working on pc386 & a couple more in powerpc. Links to accepted commits can be seen here

As a high school and a college student now, I've always been fascinated by Facebook and it's culture. Their core philosophy goes by the words "Move Fast and Break Things". The meaning behind this slogan is not to cause havoc, but instead to encourage engineers to build more things and learn faster. The idea is that if you never break anything, you’re probably not moving fast enough. Hence, you're not learning.

Although I deeply understand this quote, and feel motivated by it, I seem to be taking the above lines a bit literally while working on my GSOC Project. I feel a need to move fast with my work, sending patches in an accelerated sequence again & again until I've corrected them and they're accepted. During this process, Joel & Sebastian seem to be always behind me pointing me towards the things that I've actually broken along my way, instead of improving them. I've been asked to move forward carefully reminding me to check my work for quality, rather than quantity, and I appreciate this suggestion.

What I would like to summarize with this post is that, I'm maybe breaking more stuff at a time, rather than fixing it, but I'm definitely fulfilling the most important aspect of this above philosophy. That is, I am learning. And I am improving my work, along with maintaining my speed.

No comments:

Post a Comment