building a contribution culture @cloudwatt
TRANSCRIPT
Building a Contribution Culture @Cloudwatt
Régis Allegre
VP Software Engineering
@HappyKing
Loic Dachary
Software Craftsman
OpenStack Summit Atlanta – May 13th 2014
- Sovereign Cloud provider, created to deliver competitive
IAAS Services to French & European companies
- Joint Venture (Orange, Thales & French Gov.)
- Security, privacy and resiliency compliant with the French &
European legislation and regulation requirements
- A Free Software DNA from inception
- One of the largest OpenStack deployment in EU
Building a European Leader
on Free Software foundations
Confidentiel - Préliminaire 3
Why did we chose to contribute?
1. Solution Mastery
2. Influence on our Architectural Foundations
3. Sovereignty: Developing our own Competency &
Expertise
1. Access to world-class technical coaching
2. Access leading edge engineering practices & tools
3. Boosts attractiveness for the « right » developers
4. Leverage opportunity on new developments
The Necessity
The Bonus
Ala Rezmerita
Cédric Soulas
Jordan Pittier
Nassim Babaci
Sahid Ferdjaoui
Edouard Thuleau
Loic Dachary
Christophe Courtaut
Yves-Gwenael Bourhis
Cloudwatt as a Contributor
9
Contributions Other Activities
Source: www.stackalytics.com
Team
18th contributor (Icehouse)
25 Proposed Blueprints
5 Completed Blueprints
4th contributorProjects
Our Options
10
Feature is submitted, and gets through the acceptance
process
Feature is implemented an external component, integrated
through APIs with the core project.
Upstream
External Component
Downstream Feature is implemented as a fork from the regular
component
11
Upstream
Change
External Component
Downstream
Change
Contribution will be reviewed,
improved and validated
Maintained with the product « by
construct »
Can be developed by others
(leverage)
Feature remains differentiating
Can be implemented faster (you
own the lifecycle)
Feature remains differentiating
Can be implemented faster (you own the
lifecycle)
Feature is not differentiating
Can be long to pass through
Can be rejected
Change needs to be reapplied
(recurring cost)
Can be incompatible with
future choices
Feature will not be improved
by the review process
Subject to API evolutions: The
component will need to be
maintained forever
Significant architectural impact
(complexity, integration,
performance)
Pros Cons
Our Options
Hard questions
12
You’re tiny fish. Do you really expect to influence anything when the
big guys get involved?
When insert_big_company_name contributes, it’s a rounding error
on their P&L. Can you afford it as a small scale player?
How do you build your company culture, when your developers
loyalty go primarily to the open source project?
That’s a lot of people you’re flying to the Summit. Is this
reasonable?