The focus this week has been on our FOSS project rather than individual assignments. There is not much to say here that is not on my group's wiki page. I can briefly mention the work that I have done individually, but four separate blog posts would just be a waste of everyone's time. This post is a condensed version of what I have been working on over the past week.
Release Candidate 2 (RC2) was released on March 2 and further refines features that were added in RC1. All of the recent changes can be tracked here on XBMC's github page. RC2 quickly fixed a few major bugs, such as video lag, that were present in RC1. It is clear that the developers rapidly respond to problems that are readily defined in the bug tracker and on the forums. RC1 and RC2 were released only a week apart. The first order of business during our first group meeting after spring break was to catch up and compile the latest release from source. A weekly release schedule can eat up a lot of our time: it takes over an hour to compile on my poor little netbook. Speaking of my netbook, Ubuntu 12.04 is running nicely, and all of my previous problems appear to be fixed with this latest update. The newest kernel includes a fix for the back light problems that most netbook users are having.
It is crazy to think about how much XBMC has changed since we first started working on this project. When I first mentioned XBMC to my group earlier this semester, I was using Dharma on my desktop. Now, we are three beta releases and two release candidates into Eden. The pace is pretty hectic, and any contributions in this project are welcome to take some of the stress off of the developers. Matt and I replied to the Xbox 360 controller bug report, and offered to help fix the problem now that it has been discovered by the developers. CrystalP asked for us to test the changes and submitted a pull request (PR) on github. Git has a high learning curve. My experience with SVN last semester was pretty straight forward, but Matt and I had some trouble figuring out how to actually implement a PR in Git. There are a couple of Stackoverflow questions and various blogs that suggest possible solutions, but there are so many conflicting answers. Cherry-pick seems to be the best command to pull in a patch from a single commit. The process is very convoluted, and, unfortunately for our group's contribution (fortunately for the entire XBMC community), CrystalP went ahead and implemented the PR. At least we contributed to the conversation and gave the bug the push it needed to be fixed for Eden.
Jason's bug was also fixed by a developer while he was looking though the code for possible solutions. I let David borrow my netbook to test more LIRC bug reports, but these reports seem to be very hardware specific. He has not had much luck replicating bugs that deal with LIRC. It is difficult to keep up with such a fast-paced community with everything else that is going on this semester. We have decided to continue working on any bugs that come our way, but we need to refocus some of our effort to areas where we can contribute more. In our next meeting, we will come up with a schedule for our poster and find a couple of important pages on XBMC's wiki that need to be updated. Much of their wiki was designed for Dharma, and they desperately need some updates in order to be ready for Eden's release. From now on, I will focus my blog posts that are not dedicated to individual assignments on possible bug fixes to discuss with my group or specific pages on the wiki that needs to be updated. I hope this condensed post is more informative then three or four scattered posts.
No comments:
Post a Comment