From Fedora Project Wiki

The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

Roll Call

  • Present: Christopher Aillon, Bill Nottingham, Dimitris Glezos, Mike McGrath, Dennis Gilmore, Matt Domsch, Paul Frields
  • Regrets: Josh Boyer, John Poelstra, Tom 'spot' Callaway

Last meeting

Meeting:Board_meeting_2009-10-01

Proposed Agenda

Trademark license agreement update

  • Current status
    • Pam Chestek did collaborative editing on the wiki with input from FAB
    • Final version has been created
    • Multiple recipients in queue to receive it
    • Door is open for new signers in the future
    • We will also offer to current holders so that they have the option of switching to the new agreement

Commercial non-software goods license

Meeting:Board_meeting_2009-09-10#t10_Sep_13:25

  • Current status report via Paul
    • No time in Comm-Arch team schedule for a Finance meeting
    • Paul re-tickled Max, we will set up a meeting to find out whether we can set up a receiving fund for Fedora

Target audience for distribution

User:Pfrields/FAB_roundup

  • Paul: The context is "Where do we want to be"
  • Need to have more definitive list of the things we expect the target audience to be able to do with the Fedora distribution
    • Installation/upgrade
    • Specific core tasks (reboot, connect to Internet, system update, browse, IM/IRC)
  • and what the functional tolerance is -- meaning what is acceptable performance for each of these
    • Time might be one criterion, errors/fallback might be another
    • This helps us make better release criteria, determine blockers, etc.
  • mmcgrath: Some people think Fedora is for everyone, some don't.
  • Chris - The conversation may actually not be about who Fedora is for, but rather people just being unhappy with the updates.
  • General agreement that releases are not known for quality, updates after a release get even worse.
  • Paul: Probably ad-hoc decisions about updates, blockers, prioritization
    • Example - we provide help via IRC but a default install didn't include a IRC client
    • Jesse posted his idea for an unfrozen rawhide and better-managed current release target repo
      • Board agrees: GO FORTH AND DO.
    • Paul: We should set the audience before setting the process
    • caillon: but if we don't fix the process, our target will suffer no matter who they are
      • Paul agrees, you can't do one and not the other
    • Paul: Our vision should be: "to better fit the needs of *this* audience"
      • People voluntarily switching to Linux, not really "my aunt Tessie"
      • People who are not necessarily hackers, but are familiar with computers
      • People who are likely to fix something that is not working (or at least collaborate or report when it's not working)
      • List of tasks one can do?
        • web browsing, email, office productivity, graphic arts, publishing, audio listeners, web serving, collaboration & communication, software developers
    • There may be network benefits of this approach - by targeting these, we may be more usable for other cases, even if they're not our goal.
    • dimitris: Perceived lack is around polish. We may not be missing broad targets horribly, but there's a lot of fine tuning that needs to be done better.
    • mmcgrath: A big thing we're missing is self-control with updates.
    • mdomsch: (channeling skvidal) You as a Fedora packager essentially have root access on millions of systems. Treat them with appropriate care
    • mmcgrath: have FESCo and/or QA come up with a mandated policy around stable release updates.
    • caillon: also need to make sure that people follow those policies; see also https://www.redhat.com/archives/fedora-advisory-board/2009-October/msg00100.html
  • Decision item needed.
    • What is our target audience? (see above)
    • The Board will now pursue goals for process changes that make quality higher for that target
  • mdomsch: Board focus on quality instead of quantity of packages
    • mmcgrath: how can we measure quality?
      • How many update breakages?
      • How many updates overall? (This is more symptomatic than root-cause related)
    • This is an area where we can learn from others.
    • critical path packages need push approval for stable releases. Perhaps we don't have the resources in place to do all packages.
    • dimitris: If we increase visibility when a breakage happens, it's less likely people will be careless. Are we giving enough visibility to breakage, or do we just fix it quietly?
      • Sheriffs will help in this visibility.
    • mmcgrath: How do we prevent "Daddy says no, I'll ask Mommy"
      • caillon: With sherrifs, this has not been a big problem. All nos get logged, so that should be referenced before saying yes.
  • ACTION: Paul will follow note publishing with a summary to FAB of the agreed points:
    • Target audience statement from above, which represents broadest consensus, to be further specified collaboratively
    • Board will set update discipline goals and look to FESCo to help design/implement
    • Board concurs with Jesse Keating's expanded explanation of the "Unfozen Rawhide" proposal and will work to make it happen fully in the F13 cycle.

New Business

Elections

  • mdomsch looking for someone who might like to take up Fedora election coordination work
    • Which groups are up? Board 1/2, FESCo 1/2, Ambassadors (all?), F-13 name... who else?
    • What events need to be held?
    • When will they be held?
      • start elections after FUDCon, which closes Dec. 7
      • use FUDCon for an additional in-person town hall meeting
  • ACTION: MDomsch will kick this off on FAB, and look for existing and new volunteers to drive it

Next meeting

  • AGREED: 2009-10-29 UTC 1600