Skip to main content

Notice: this Wiki will be going read only early in 2024 and edits will no longer be possible. Please see: https://gitlab.eclipse.org/eclipsefdn/helpdesk/-/wikis/Wiki-shutdown-plan for the plan.

Jump to: navigation, search

Project Management Infrastructure Redesign 2011

Revision as of 23:16, 6 September 2011 by Wayne.eclipse.org (Talk | contribs) (New Project Proposal)

This effort is being tracked by Bug 243223.

Problem Statement

Current infrastructure (i.e. The Developer Portal) is inadequate.

Multiple, separate data sources.

Portal is separate from the the resources being managed. Requires a context switch to use. Most committers have difficulty (or outright refuse) to make that context switch.

Some management tasks are spread out. Specifying a description for a project, for example, requires that an HTML file be created in the project directory (requires CVS check-in), and then the specification of a URL in the portal. Very difficult to maintain. Very separated from where and how the description is used. As a result, descriptions tend to be poorly specified, and maintained.

Too much information is not included in or managed by the portal. Project proposals, review documentation, IP logs, are all separate.

Technology Choices

Project management is essentially a document-management and workflow problem. Several solutions exist in this area.

The Eclipse Foundation currently uses Drupal for Eclipse Marketplace, Eclipse Live, and the EclipseCon Website. Several Eclipse Foundation employees are already well-versed in Drupal development, and finding temporary resources with the necessary skills in the local area should be relatively easy and cost-effective. Drupal is based on PHP, a language that is known to most of the Eclipse Foundation staff, and is currently in wide deployment by the Eclipse Foundation.

Perhaps one of the features that weighs most heavily in Drupal's favour is the size of the community behind it (which measures in the hundreds of thousands) and the hundreds (perhaps thousands) of plugins that are available to extend it. The availability of plugins, combined with the relative ease with which Drupal can be extended means that the overall amount of custom code that needs to be maintained should be relatively small (as compared to other solutions that may require more customization).

There are several other options that have been considered, including a handful of Eclipse-based solutions (which would allow us to "eat our own dogfood"). After careful consideration, however, we have determined that we do not have the resources to implement these solutions.

Technology comparison
  Pros Cons
Drupal
  • Skilled resources within existing Eclipse Foundation staff
  • Large number of skilled resources available
  • Very large community of developers, adopters, and users.
  • Deployed by 1,000s of organizations
  • Hundreds of plug-ins available to leverage in favour of writing custom code
    • Integration with Facebook, Twitter, etc.
    • Access information via RESTful webservices
  • Build-in (no-brainer) database support
  • Integrated Development tools + Eclipse/PDT
  • Existing IT Infrastructure support
  • Drupal-specific data structures and formats.
  • Not dogfooding; perception in the community
Skalli
  • Java based: Some Java skills on staff
  • "Dogfooding" solution
  • Eclipse-based development tools available
  • Leverage existing Eclipse technologies (EMF, workflow projects, SOA, EclipseLink, etc.)
  • Project developers have expressed interest in implementing some Eclipse Foundation processes (though with no specific commitments)
  • Opportunity to fully control the database structure
  • New project, very small community
  • Limited availability of skilled resources
    • Ramp up time for new developers is relatively long.
    • Good quality Java-savvy developers are relatively difficult to find.
    • Virtually impossible to find developers familiar with Skalli itself
  • Integration opportunities with existing Eclipse technologies remain largely unexplored
  • No existing IT Infrastructure support
Apricot
  • Java based: Some Java skills on staff
  • "Dogfooding" solution
  • High-quality, established project with several high-volume consumers
  • Eclipse-based development tools available
  • Leverage existing Eclipse technologies (EMF, workflow projects, SOA, EclipseLink, etc.)
  • Community size is unknown
  • Unknown availability of skilled resources
    • Assumed to be no skilled Apricot developers available in the local area.
    • Ramp up time for new developers is relatively long.
    • Good quality Java-savvy developers are relatively difficult to find.
  • Availability and usefulness of extensions is unknown
  • No existing IT Infrastructure support

Roles

  • EMO(ED) - EMO Executive Directory (i.e. Mike)
  • EMO(PM) - EMO Project Manager (i.e. Wayne)
  • EMO(LC) - EMO Legal Council (i.e. Janet)
  • EMO(IP) - EMO Intellectual Property Team
  • Proposer
  • Committer

Communication Channels

The communication channels used by the system are expected to change with time as technologies and communities evolve. Favoured technologies of the day are Twitter, RSS, and the eclipse.proposals forum.

Core Requirements

  • New Project Proposal
  • Proposal becomes a Project
  • Elect a Committer
  • Project Release


Proposal Becomes a Project

  • New project is created
  • Some information copied from the proposal
    • Id, short/long name
    • Description
    • Scope
  • New users are created for committers as necessary
  • Committer/Project Lead/Mentor relationships captured
    • Project id
    • User id
    • Active Date
    • Committer relationships remain inactive until approved by EMO(IP)
  • A project retains linkage to the proposal.
  • A project is made active when approved by the EMO(PM), Webmaster, and EMO(IP)
  • Project's Active Date is recorded

Other

  • A project's metadata can be edited by the project lead, or any active project committer.
  • Any change to project metadata is tracked (including the date of the change and the identity of the individual who made the change).
  • Project id and scope cannot be edited

Provisioning Phase

During the provisioning phase, project resources are allocated. This includes such things as space on the various servers, creation of project and committer records, and so-forth.

Committers

Nominate a Committer

Nominate a committer to an existing project.

Workflow:

  1. Existing committer nominates a new committer, specifying:
    • name, email address
    • Paragraph describing merit of nomination
  2. Existing committers vote
    • Election ends successfully if all committers vote positively (i.e. goto 4)
  3. After one week, results are checked
    • Election ends successfully if there are at least three positive votes, with no negative votes (i.e. goto 4)
    • Election fails otherwise (i.e. exit)
  4. Notify PMC of vote results, request PMC approval
    • If approved, continue
    • If disapproved, election fails (i.e. exit)
  5. Committer provisioning commences

Provision a Committer

Workflow:

  1. EMO(IP) is informed of the need to provision committer, and is asked for approval
    • EMO(IP) works with the project, committer, and PMC to resolve any issues.
    • If approved, continue
    • If disapproved, notify project, committer, PMC, remove the request and exit
  2. Committer record is created
  3. Webmaster is informed of the need to configure backend access (i.e. add new committer to appropriate UNIX group)
  4. Webmaster indicates that backend access is established
  5. New committer, project, and PMC is informed of successful completion of committer provisioning

Back to the top