While I have been a Fedora user for years, I never really got involved with the community until I was recently hired by Red Hat to work on Fedora QA. I think that the ideal situation would have left a little more time for me to come up to speed with at least some parts of the Fedora project but how often is life ideal? Either way, it was a great opportunity to meet some of the community in person instead of just through email and IRC.
After arriving in Phoenix and making my way to the hotel in Tempe, I went down to the hospitality suite to take part in some games. I didn't end up playing any games but I did meet a lot of new people - I just wish I was better with names than I am. In particular, I did spend quite a bit of time talking to Lucas and Susmit. I also met my other North America based co-worker, Adam.
This was my first experience with a BarCamp style conference and I was a little frustrated at first because the schedule wasn't pre-determined. Once we got started with the pitches and voting, I understood more about the advantages that BarCamp has to offer but there were still scheduling conflicts for the talks I was most interested in. I suppose that is an eventuality since I had at least some interest in most of the presentations that were selected.
Fedora on ARM and Fedora on OLPC
I've been interested in doing more hacking on ARM for a low power NAS as of late, so I was really interested in learning more about Linux and specifically Fedora on ARM. I learned a lot about the differences between the different ARM architectures and how those differences make a unified ARM distribution difficult at best. I also got wind of some new ARM devices that I didn't know about - the Toshiba AC100 and the PandaBoard. I'm hoping to get a PandaBoard to hack on once I get some other projects under way.
I'm going to be spending a lot of time working on AutoQA, so of course I went to that presentation :) To be honest, I didn't really understand everything that was going on due to my lack of background knowledge (koji, bodhi, packaging processes etc.) but I like the fact that so many people were interested in testing automation. Hopefully we can keep community interest and get more participation.
Transcribing via IRC
There wasn't a talk on transcription but after the AutoQA presentation, there was request for people to start transcribing the presentations on IRC. For some reason, I thought that it would be a good idea to give this a try. I managed to transcribe most of the presentations I attended but I never really realized how slow my typing was until I started trying to keep up with the conversations in these presentations. I think that I was able to record a decent chunk of the information, though. As a side note, MeetBot is great if you know that it exists - I didn't until the second day of presentations.
Different species of Python
After a hiatus of longer than I would care to admit, I'm getting back into using python on a regular basis. I was vaguely familiar with some of the non-cpython variations but wanted to learn more, especially about pypy. I can't say that I walked out of the presentation with a complete understanding of the Python species covered in the talk (damn my inability to learn at an infinitely high rate) but I do know a lot more about Python bytecode and how it differs between versions and implementations of Python.
Meet the Anaconda team
Who doesn't like installers? My attendance here was motivated partially by the fact that Fedora QA works with the Anaconda team quite a bit and I'm interested in seeing where Anaconda is planning to go in the near future. There was a lot of talk about the coming changes to PyGTK or rather the fact that it will be going away in favor of PyGobject. It will be interesting to see how the Anaconda UI changes as it is rewritten for PyGobject. There were also some good examples of the new storage testing interface for Anaconda and how it can help developers reproduce storage related bugs without needing a replica of the problematic system.
Using the SELinux Sandbox
SELinux is something that I have fought with on several occasions in the past and while the default answer for systems behind a firewall seems to be "just turn it off already", I'd like to be able to keep it on and try to do so for the systems that I have control over. I never really thought about other uses for SELinux but I now know that you can use it to isolate applications like Firefox if you need/want to visit a site that you don't trust.
I'm not always a fan of pool or bowling but FUDPub ended up being a lot of fun. Where else are you going to find a penguin playing pool? I got talked into playing some pool myself and after getting soundly beaten several times, I managed to find some people who weren't much better than I was and had a couple of closer games. We were eventually kicked out of the union when it closed and I learned something - don't follow drunk people around a city that they don't know.
AppStream and 10ish Things You Didn't Know about Yum
I certainly didn't know that you can do a decent amount of searching in yum, I've always just used stuff like 'yum search blah | grep foo' to better parse output. I'm going to have to play around with that in F15. AppStream is an intriguing concept that, if implemented fully, would allow Linux users of any distribution to install a common set of packages instead of the current system of only distro-specific packages. It sounds like the work on this is pretty young but I'll be interested to see where it goes.
Writing SELinux Policy
Again with my interest in SELinux :) I'm going to have to play with this a lot more before I'm able to write SELinux policies but I definitely know more about types, domains and SELinux tools than I did.
At first, Matahari sounded a lot like puppet but at the end of the day, it is a remote interface to a wide variety of machine types, not just configuration management. This could be used with (and plugged in to) systems like Spacewalk or puppet, but would also allow for better collection of statistics and a better sense of whether the OS inside a VM is actually running instead of assuming that the OS is running if the VM is.
As they are only 5 minutes, it is hard to cover a whole lot of material but I got several take aways (mostly notes for more investigation on my part) from the lightning talks:
I wasn't able to help out a whole lot, but I did attempt to do some packaging for Fedora Medical. My attempts to complete a package were thwarted by licenses (a pandora's box of proprietary and open source licenses, very poorly documented). I also spent some time talking with lmr and jlaska about autoqa and autotest.
Overall, I was able to meet more of the community and learn a lot. I skipped most of the hackfests this year since my lack of prior experience would have limited my ability to make forward progress. I'm looking forward to the next FUDCon and participating in the hackfests
Note: Edited to fix broken formatting after initial posting