From Mageia wiki
Jump to: navigation, search


FOSDEM 2011 presentation docs drafts

We will have two slots:


  • 10:30 (30 min) in H.1302 (CrossDistro devroom) for a technical focus on the project
  • 16:30 (30 min) in H.2214 (LibreOffice devroom) for an organizational focus

So we will have two presentations. Each should hold at least a common trunk, presenting the project largely (5/10 minutes), followed by a specific summary (all the rest), followed by a Q&A session (5 minutes).


Those talks will be held by misc and ennael for the most part (if only).


Both talks are introduced as:


The difficulty of forking, and why you should not do it, yet prepare for it

Forking an open source project is never an easy task, both on a social and a technical level.


Mageia, much like LibreOffice, is a substantial, and quite large, project born from a fork, facing several issues regarding community and infrastructure.


The lessons learned from forking can be generalized with regards to any software project.


The talk will explain the difficulties inherent in forking a large, distributed, open source project (namely a Linux distribution), the hurdles, issues, and potential problems that such a project will face; how you can address these issues and fix them, and plan to prevent them in the future if and/or when a need to fork occurs again in the future.


TODO

  • record the notes _in_ the slides (for notes)
  • list the generally accepted ideas on forks and break them progressively
  • blog post on packaging/websvn/?
  • hours and points of appointment
    1. packaging ruby (shikamaru, misc, rda)
    2. General Assembly - 14:30 on Saturday ?
    3. db for bs (misc, blino, nanar, pterjan)

Common trunk, from the start

Those are the major points of the presentation. Slides on their way.


Mageia

What?

  • (not only) a Linux distribution, forked from Mandriva Linux/Cooker
  • an international community project, with a meritocratic governance
  • a not-for-profit association

Why?

  • people first, most valuable asset of a project (users & builders)
    • trust
    • goodwill
    • plans
    • moving the control of the project into the sole, controlled hands of the community: learn the technology & drive it
  • because we can
  • reveals new options of what is possible
  • because we want to
    • great 12 years track technology & experience, from the early days of MandrakeLinux
  • because that is a great time to do it

What for?

  • distribute more work through the community, multiple experiences/skills
  • define and experiment new, sustainable economic construction around it, respecting the project's core values
  • get a better understanding of users and usages
  • grow the culture of learning through the project, of so many connected topics: software development, free software, creation, personal data & uses, community & sharing economics
  • grow the culture of the technology/economics being trusted tools for people growth in their daily lives, decisions, actions
    • requires fun, goodwill, trust
    • requires listening, empathy, zen
    • requires problem solving, positive attitude
  • get to a better hardware integration
  • not only be another Linux distribution
  • http://mageia.org/en/about/values/

For who?

  • contributors
  • developers, hackers, students
  • everyday users (see above, see below)
  • schools, admins, labs, orgs
  • all disciplines encouraged

Fork

  • launched by a team of 17 ex-employees & long-time, experimented contributors at various levels
  • wonderful, unanticipated feedback & donations
  • about 600 volunteers cheered in
  • about 50 very active people right now
  • 15 teams
    • distro/dev, packaging, triage, i18n, sysadmin, web, artwork, marcom, QA
    • pending: Doc, legal, forums/ml/irc, local communities, UX/Design, Hardware

Governance

  • community (needs, ideas, means, people)
  • teams (reflexion, production in specific areas); has peers & elected leader & representative
  • council (driving), made of representatives
  • board (direction), elected among past representatives (founders the first year)
  • association (legal structure), lead by the board, composed of all active past representatives

Roadmap

In September, we had 4 milestones/goals (aka, “define success”) in view; as per the project Mageia as a whole, the goal is to fork a totally independant project, on technological and organisational grounds:


  1. “get things started”: success (expected to take 1 month, took 2/3 in the end)
    1. setup the legal and trusted governance structures, start the whole thing, gather teams, put everyone to work
  3. “get a working factory”: success (expected 3 months, took 5)
    1. ⇒ we are here today
    2. we have a clean, independant, owned and mastered technical platform to go ahead with the technology and the project
    3. finally, we took 4 months to be there (alpha 1 released a few days ago) « not done yet
    4. so DO NOT EXPECT our alpha 1 to be nice-looking platform; if you're not an active contributor, don't even test it!
    5. no particular innovation for now (git, packaging, release cycle, bs)
    6. that was planned, goal was to migrate the platform, clean it up, improve a bit the tools
  5. “release a stable distribution”: (expected 6 months at first, will most likely take 9)
    1. still pending, yet on track and making continuous, acceptable progress toward that realistic, and achievable target
    2. have a stable, public final release of Mageia 1
    3. have a working governance/contribution structure
    4. have elected team leaders
  7. “brainstorm your community in 9 months”: still pending
    1. brainstorm the project roadmap
    2. have companies/universities join/gather around the platform
  • [now, to the specific part of this talk]
  • closing:
    • it's fun!
    • it's started!
    • you're welcome to join!
    • see you in one year
    • in the meantime, if you want to help: mageia.org

Technical focus

Here we will talk more of the technical facets of the project; what we have done and what we expect to do.


Forking process

  • importing & cleaning source code and packages: checking and setting licenses rights, trademarks, removing unlicensed artwork
  • ensured everything is under a free software license
  • facilitating a white mark derivative or even a new fork
  • cleanup attic stuff
  • “state of the kitchen”

bootstrapping the distribution

  • what is it?
  • the distribution had never been thought for bootstrapping
  • it happens not very often, it's not well documented
  • however an easily bootstrapable distribution makes for:
    • a quality check
    • an easier subsequent fork
    • easier ports to new architectures
  • we will need/have a permanent tool to check and indicate how well the distribution is bootstrapable at any point in time (part of the continuous integration effort in the whole Mageia build system)

better documentation for packagers and developers

  • bring mageia as a development/experimentation platform of choice

build system

  • we had a choice to make in a short timeframe:
    • not too long a learning curve, so that we were operational in 3 months (it happened in 4 actually)
    • very small infrastructure to start with (about 3 servers for bs)
    • why not fedora bs? too linked to yum
    • why not opensuse bs? not sized to our infra (needs a lot of servers, uses VMs, has its own version management, own workflow)
    • so use and improve (quite a lot already) existing one

Systems Administration (our sysadmin team)

  • Progress on becoming self-hosting
    1. Mageia infrastructure is approximately 90% self-hosted at this time on the Mageia.org domain
      1. Mailing lists / List managers (Mailman)
      2. PostgreSQL
        1. bugzilla
        2. sympa
        3. epoll
        4. mga::mirror
        5. transifex
      4. MySQL
        1. package database
        2. Mageia blog
      6. Puppet/SVN
        1. Publicly accessible
        2. Web interface is still pending
      8. Main Website
    3. Forums and forum gateway integration with email lists pending
    4. Migration of Mageia Wiki to permanent hosting pending
    5. Download Mirrors (HTTP / FTP / Bittorent)
      1. pending

In case those are raised in the discussion, prepare:


Organisation focus

Here we are going to talk more about the organization of the project, its governance and goals.


  • we have to expect an enthusiasm “it is necessary to change everything, this, that” which we have to channel in a constructive flow of energy
  • difficulty in maintaining motivated the contributors when there is not yet things released and when everyone is not yet busy by working on the project (graph about the traffic on the -discuss mailing list and on the webiste)
  • what sets Mageia aside?
    • it's a set of things
    • core values
    • governance & project learning organisation
    • not Linux as the only goal and target
    • ecosystem
  • governance focus
  • project roadmap focus?
  • teams renewal; need for experts/mentors; need for juniors
  • new ideas

Longer roadmap

  • turn values into action
  • distribution targets
  • complementary projects
  • ecosystem

Funds/Donations

  • what was received
  • what was used
  • what are the projections/needs for the coming year

Ecosystem

  •  ?

Questions to prepare if asked:

  • relationship with mdv? nothing specific. we will collaborate with all open source projects

Draft documents

Both share a first common trunk to present the project as a whole. These are work-document, not final ones.


Retrieved from "https://wiki.mageia.org/mw-en/index.php?title=Archive:Fosdem_2011_presentation_draft&oldid=57776"