AMO:Roadmap/2011

From MozillaWiki
Jump to: navigation, search


Addonsicon.png Add-ons 2011 Roadmap
Owner: Justin Scott Updated: 2011-02-4
In 2011, our key initiatives are focused on platform, performance, and penetration. These are the 3 most important things for us to focus on, and our goal is to make the following statements true by the end of the year. PLATFORM: Firefox is the most attractive platform for browser add-on development, PERFORMANCE: Add-ons don't degrade Firefox performance, PENETRATION: 60% of Firefox users use extensions.


PLATFORM: Firefox is the most attractive platform for browser add-on development

In the last year, the browser add-ons landscape has been turned upside-down with 3 new entrants: Chrome, Safari, and Opera. All have restartless, easy-to-develop platforms and developers are flocking to them. We need to re-establish our dominance in add-ons and, hopefully, web/app platforms in general.

Where we are right now

Our XUL platform is extremely powerful but seen as outdated and hard to learn. Add-on SDK is making good progress and promises restartless, process-separated add-ons that are very easy to learn and write.

Where we want to be

  • Developers have all the tools, documentation, and support they need to build great add-ons
  • It's easy to get off the ground quickly with the SDK
  • There's a clear path to monetization (or voluntary contributions) for interested developers
  • Firefox and our platform update quickly to fix bugs and provide new features without introducing compatibility hassles
  • Add-ons want to be hosted in our gallery because of its unmatched add-on distribution and discoverability channels
  • Add-on review process is seen as a helpful validation rather than an annoying burden

How we'll get there

Q# Initiative Difficulty <thead> </thead> <tbody> </tbody>
style="background:#efefef" Dependencies/Resources
1.1
Solution to frequent compatibility updates
Hard
Add-ons Manager, AMO, Add-on DevRel teams
-
1.2
Rewrite Editor Tools to improve review experience
Medium
AMO, Add-on DevRel teams
-
1.3
Switch Contributions to PayPal in-context flow
Hard
AMO team
-
2.1
Launch marketplace to support freemium add-ons
Hard
AMO, Marketing, PR, Add-on DevRel, Legal teams
-
3.1
In-app purchases support in marketplace
Hard
AMO team
-
{Myk's Add-on SDK items will go here}
}

Measuring progress

KPI Start <thead> </thead> <tbody> </tbody>
style="background:#efefef" Goal
New extensions submitted to AMO per month
302
600
-
New extension developers on AMO per month
204
350
-
% of new add-ons using the SDK
3%
60%
-
New add-ons created using the Add-on Builder per month
???
1000?
-
Average review time for new add-ons
TBD
TBD
}

PERFORMANCE: Add-ons don't degrade Firefox performance

Add-ons are one of biggest reasons people use Firefox, but their performance impact is also one of the reasons people leave, whether they realize it or not. The worst offenders are add-ons not hosted on AMO and installed via third parties, but even hosted add-ons share some of the blame. It's time to get serious about add-on performance impacts.

Where we are right now

The Automation team can manually run startup (Ts) tests for the most popular add-ons, and AMO is ready to show performance warnings on those add-ons alerting users to the slowness.

Where we want to be

  • Establish a measurable metric for overall add-on performance that we drive downward over the year
  • Performance documentation and best practices are provided for developers to follow
  • Policies are set in place and extreme violators are removed from AMO and/or blocklisted
  • Developers can test their add-on's performance impact with tools BEFORE exposing to end users
  • Users are aware of the performance impacts of their add-ons and must consciously choose to install a performance degrading add-on
  • Adoption of the SDK platform grows and brings performance improvements with it

How we'll get there

Q# Initiative Difficulty Dependencies/Resources
1.1 Establish performance policies for hosted and non-hosted add-ons Low Add-on DevRel
1.2 Gather data on add-on performance in the wild and reach out/blocklist Medium 593743, Add-on DevRel
1.3 Automate performance tests for top 90% of AMO add-ons High Automation team
2.1 Expose performance information in Firefox Medium Add-ons Manager team
2.2 Provide on-demand performance analyzer tool High Automation & AMO team
2.3 Launch performance campaign directed at developers Medium Add-on DevRel, 1.3, 2.2
3.1 Expand performance tests to include page load time (Tp) High Automation & Platform teams

Measuring progress

KPI Start <thead> </thead> <tbody> </tbody>
style="background:#efefef" Goal
TBD (see 1.2), but probably average start-up time per installed add-on
unknown
TBD
-
Percentage of add-ons over "warning" threshold
unknown
TBD
}

PENETRATION: 60% of Firefox users use extensions

Users who install add-ons are more likely to stick with Firefox and become loyal fans and evangelists. Rather than trying to increase the number of add-ons users have installed, we should try to increase the number of users with add-ons.

Where we are right now

At last measure in October 2010, 40% of all Firefox users had at least one extension or theme installed. Firefox 4 provides a way for us to consistently and reliably measure this percentage, and after release we can use that as the baseline.

Where we want to be

  • 60% of Firefox users have at least one extension (not Persona, theme, etc.) installed
  • We don't count third-party-installed add-ons such as Java Console

How we'll get there

Q# Initiative Difficulty <thead> </thead> <tbody> </tbody>
style="background:#efefef" Dependencies/Resources
1.1
Finish Discovery Pane implementation for Firefox 4
Low
AMO team
-
1.2
Begin locale-specific featured add-ons
Low
Marketing
-
1.3
Mobile-friendly gallery pages
Medium
AMO team
-
1.4
Revamped homepage and details page
Medium
AMO team
-
1.5
Add-on "getting started" page with video
Medium
Marketing
-
4.1
Localization campaign to localize more add-ons and metadata
High
AMO, l10n teams
}

Measuring progress

KPI Start <thead> </thead> <tbody> </tbody>
style="background:#efefef" Goal
Firefox (desktop) users with an extension installed
40%
60%
-
Firefox (mobile) users with an extension installed
TBD
TBD
}

Time Vampires

The projects below don't further our goals above, but will still need to be done as part of infrastructure/platform maintenance.

Q# Project Difficulty <thead> </thead> <tbody> </tbody>
style="background:#efefef" Dependencies/Resources
3
GetPersonas.com integration with AMO
High
AMO & Personas teams
-
3-4
Finish migrating pages to Django and decommission PHP
Medium
AMO team
}