if by culture you mean fat people.

February 4, 2007

FudCON and the hackfests this weekend were fantastic. I was skeptical about the barcamp format at first but it appears to have worked out just great. I made a short list of things I have to play with as a result of what I heard or read this weekend:

  • xen live migration
  • libvirt
  • pull-from-git and build a kernel in a single command
  • smolt work:
    • better hardare.Hardware abstraction layer
    • separation of public/non-unique and private/unique data
    • smoon (smolt satellite)

I’m fairly pleased when a number of issues I’ve been dealing with have been tackled by other folks. Now I can just jump on their solutions or enhance them and work from there.

The yum hackfest days went far better than I thought they could have. We had a number of high priority issues and we knocked out all of them. We’ve gotten some very preliminary tests done and it looks great. Here’s what we did:

  • made createrepo able to generate the sqlite db’s on the repository side
  • made yum look for those first and grab them instead of the xml (in some very simple and non-definitive tests this affords a 28% speedup versus parsing the xml into dbs)
  • brought in dlutter’s patches to make yum use the database backend a bit more intelligently and significantly speeds up package-name matching operations (yum install, erase, update with pkgs specified)
  • grabbed the header-less-dep-solve mechanism from anaconda and put it in place for some early testing. It is NOT ready for regular use yet (for example, erasure deps don’t work, you know, at all) but we should be able to get it brushed up and happy before T2, I hope.
  • media handling is implemented in yum in some fairly simple code from Jeremy.

What was most amazing was how much we were able to sit down and just do. I hope to put a yum 3.1.2 in the next couple of days to find out how much we broke in the process of speeding things up. Additionally, I got to speak with David Lutterkort and David Malcolm about enhancing our ability to know that yum hasn’t regressed from release to release. Dmalcolm is working on some specs for  unittests/test harness for yum.

Other results from this weekend:

I got to meet: Mike Mcgrath, Denis Gilmore, Toshio Kuratomi, Bob Jensen, Jef Spaleta, Thomas Vander-Stichele, Thorsten Leehumis, David Lutterkort, David Malcolm, Brad Smith, James Bowes and Rahul.  It was great to meet everyone in person and talk to them about what they’re doing and maybe what I can do that can help make them more possible.

The issue was raised at the board meeting and the board dinner that we may want to have a hackfest after every Fedora Test1 release. There is a little bit of a lull at that point and it’s helpful to get everyone in a room for a while to solve the issues we have (some technical, some personal)

I have to agree with the assessments of many other folks on fedora people and #fedora-fudcon. This was the best fudcon so far. A lot of momentum and we were able to keep it up throughout the weekend.

Now, if we can just keep the trips to the emergency room to a bare minimum. 🙂


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: