May 29, 2012

The next trip report is from Mark Linimon:

Recently, thanks to Foundation funding, I was able to travel to Ottawa, ON, to attend the FreeBSD 2012 Developer’s Summit and BSDCan 2012.Here are some of the highlights.

I arrived early to be able to meet informally with people before the “big rush” began.In particular, both before and during theconference, we had meetings among the people responsible for wrangling the FreeBSD.org machines, including Simon Nielsen, Brad Davis, Sean Bruno, Ben Haga, Peter Losher, myself, and others.These discussions were quite valuable to determine what we can do in both the short- and medium-term with respect to bringing new machines and functionality online.

On the first day of the DevSummit, I presided over two workshop sessions on “the state of the FreeBSD Ports Collection.”The first session was a summary of what the ports teams had accomplished in the preceding 12 months:

  • Hardware updates to the pointyhat build clusters.
  • The redports distributed tinderbox system.
  • Status of ports on FreeBSD 10.
  • Status of ports on clang.
  • A brief survey of pkgng.
  • Status of the “switchable ports compiler” task.

The second session was a discussion of plans for the rest of the year:

  • Package sets.
  • The VCS switch to svn.
  • The status of next-generation options code, optionsNG.
  • Plans for a Content Distribution Network.

These sessions seemed to be well-received.I have made notes of some of the feedback and will be incorporating them into a writeup soon.

During BSDCan itself, I presented some summary slides of the above, edited to be suitable for a more general audience.

Baptiste Daroussin’s BSDCan talk on pkgng, the next-generation package management system for FreeBSD, was the highlight of the trip for me.At 3 different times during the Q&A session, there was applause in repsonse to questions of the form “but will it do*this* too?”Congratulations to Baptiste and everyone else who hascontributed to the code and to testing it.I believe he has fulfilledthe promise that I and others saw in his ideas at BSDCan 2011.You can watch the video of his presentation on YouTube

At other times, I was able to sit down face-to-face with several people whom I had previously not met in person, who had questions or suggestions about the Ports Collection, including David Chisnall,Justin Hibbits, Tom Judge, and Andrew Pantyukhin.Each was eager to ask how they could help.

During one evening, we had a dinner session between the majority of the Ports Management Team (portmgr) who were in attendence: Thomas Abthorpe, Erwin Lansing, Beat Gaetzi, Baptiste Daroussin, and myself.This is the first time in my knowledge that there had been as many of us together in one place.Mixed in with the socializing were discussions about current issues.

I was also able to talk about documentation issues with Glen Barber, Warren Block, and others.They are bringing some new energy and newideas to our documentation team.

Among the other people I was able to have good discussions with were John Baldwin, Brooks Davis, Julien Laffaye, and Steve Wills.A lotof “oh-by-the-way” ideas tend to come up during these kinds of face- to-face discussions that simply don’t occur online, no matter howused to online communications we are.

I also made sure to remind Justin Gibbs not to forget to mention all the hardware purchases that the Foundation has recently made when he’s doing his “Foundation pitch”.These purchases have modernized the ports cluster and will be especially valuable as we start buildingpackages for pkgng in parallel with the existing technology.We would have not been able to keep up with the increased load without them.

All in all I consider this a very successful trip.A lot of good ideas were exchanged; we have several new people who seem to have a lot of energy and enthusiasm; and we were able to bring a lot of people “onto the same page”, especially with respect to the cluster administration and Ports Collection issues.

I appreciate the Foundation making it possible for me to attend this year.