So the top vote getter from the priorities poll was:
        Deploy a public defect management system

It seems there's two obvious steps to start out with, which can
be done in parallel:

1) Have the bugs.opensolaris.org team and the bugster teams
    inform us of their plans for making a open bug tracking system.
    An e-mail summary would be preferred, but if they want to make
    slides to present on a con-call, we could handle that too.

2) Decide what we expect in a public defect management system.

Once we have those we can determine if the best course of action
is to continue to push on the current bugs.os.o/bugster system or
ask the Tools community to turn the list from #2 into software
system requirements and begin evaluating candidates to replace
bugs.os.o.

Off the top of my head, for #2, I'd start with these requirements,
but expect others to think of more:

  - almost all actions available to any valid OpenSolaris
    contributor, whether Sun employee or not, including:
        - filing reports
        - updating reports
        - accepting ownership of reports

  - ability to limit visibility of reports to selected contributors
    for a limited time, to allow for keeping security vulnerabilities
    embargoed until publication, and for keeping pre-release platform
    details embargoed until release (whether those be Sun, Intel, AMD
    or any other vendor who may choose to work with the OpenSolaris
    community).

  - if not using the Sun internal bug database directly (as bugs.os.o
    does now), then the ability to have Sun employees easily clone bugs
    from the internal db on a bug-by-bug basis, to migrate data from the
    traditional Solaris bug db.   I'd rather see this as a simple tool
    anyone with access to the Sun bug db can use, so we can split the
    load of migrating the data and deciding which parts are clear to
    export among hundreds of people - give each a subcat or list of a
    few dozen bugs at a time, and use the power of the masses to make
    easier work of the impossible-to-automate process of determining
    which data is truly confidential and which isn't.

-- 
        -Alan Coopersmith-           alan.coopersmith at sun.com
         Sun Microsystems, Inc. - X Window System Engineering


Reply via email to