[oi-dev] Developer Meeting Minutes 2010/12/14

[Guido Berhoerster]

Hello,

here are the minutes from yesterdays developer meeting. The IRC
logs can be found at
http://wiki.openindiana.org/download/attachments/4883565/oi-meeting-20101214.log

-- 
Guido Berhoerster
-------------- next part --------------
Developer Meeting Dec. 14th 2010
================================

Core Developers attending:

Albert Lee (Triskelios)
Chris Ridd (chrisridd)
Guido Berh?rster (gber)
James O'Gorman (jamesog)
Yasushi TAKIAKI (kohju)
Jon Tibble (Meths)
Julian Wiesener (yofuh)
Andrzej Szeszo (aszeszo)


Topic: restructuring the project and creating a project governance model that
fits current reality
-----------------------------------------------------------------------------

Guido Berh?rster opened the discussion with two theses of the current major problems:
1) the project is missing a basic structure and opaque to outsiders which makes
   it hard for outsiders to contribute
2) there is no leadership and direction, i.e. one or several persons organizing
   the project, making decisions, and integrating everything

Chris Ridd, Piotr Jasiukajtis, and Jon Tibble suggested that current situation
with communication mainly happening on #oi-dev makes it hard for interested
parties and potential contributors to follow development and suggested that
developers make more use of the mailing lists and wiki.

Jon Tibble stated his intent to move selected content from the wiki to the main
website and to change access restrictions to the wiki allowing anyone to
register while restricting write access to a set of official pages. In addition
RFEs and support questions will need to be moved from the wiki to the
bugtracker and mailinglists.


Decisions

Due to the relatively low number of developers the plan of formally electing a
developer council will be scrapped, rather decisions regarding development
should be made by consensus.
In case a consensus cannot be reached there should be a vote (exact voting
procedure, a quorum will need to be worked out and decided upon in a later
meeting).
The initial constituency should consist of those assigned management privilege
on Redmine which includes most people working on core parts of the
distribution, namely Alasdair Lumsden, Albert Lee, Chris Ridd, Guido
Berh?rster, James O'Gorman, Yasushi Taki, and additionally Jon Tibble, Julian
Wiesener, and Andrzej Szeszo should be included due to their work on the
distribution.

The project will strive towards greater transparency, there will be a new
"Getting Involved" page directed towards potential contributors, every area of
the project should have a designated contact person, the main website will be
filled with content, and developers will better communicate what they are
working on using the mailinglists and wiki.  If possible the team leaders
should outline open tasks on this page as well.

There will regular developer meetings to make decisions regarding development
and direction.


Topic: release goals for OpenIndiana b148+1
-------------------------------------------

The main goal for b148+1 will be to replace Oracle ON with Illumos.

Parts dropped from Illumos such as Perl 5.8.4 need to be dealt with.

In case of Perl 5.8.4 Guido Berh?rster and Albert Lee argued that it might be
provided for backwards compatibility as an unsupported part of the system,
Julian Wiesener noted that many packages currently depend on it and need to be
changed, Jon Tibble and Andrzej Szeszo questioned the need to keep it around.
A decision will be postponed to a later meeting.

Note from the editor: The G11n is in a bad shape and could not be fixed in time
for the b147 and b148 releases, it will require additional adaptations for
Illumos.
Albert Lee will check with the Illumos project whether OpenIndiana can can get
any resources helping to fix it.


Decisions

b148+1 will be based on Illumos.

The G11n consolidation which has been skipped for b148+1 must be fixed and
adapted to Illumos.

A tested AI Image should be provided.

The Oracle build of XNV, SFW, and JDS which will serve as a basis for b148+1
will be determined when there is a working base system.