Support/Q42009 Goals

From MozillaWiki
Jump to: navigation, search

High-level SUMO Focus of Q4

The over-arching theme of our focus in Q4 can be summarized as:

  • Page the way for a better SUMO platform
  • Grow our SUMO community

Pave the way for a better SUMO platform

SUMO is built on a two-year-old version of TikiWiki (version 1.10, a beta version that never went GA). The immaturity of the platform has made it more complicated to add new features to SUMO and we have instead spent a great deal of our time working on performance and security improvements and bug fixes. In order to build a better foundation for SUMO, we need to upgrade and ultimately rewrite parts of the platform to ensure higher code quality and, as a result, happier SUMO developers.

The goals of Q4 are just the first steps in the effort to improve the SUMO platform. However, during this effort, we will go from a "consumer" to an active participant of the Tiki community, which will hopefully lead to a much stronger collaboration, resulting in improvements to the platform being assisted by the Tiki community as well.

The next steps would then be to upgrade to Tiki 4.1/5.1 and rewrite the bad parts of the platform to improve code quality. These steps are unlikely to happen in Q4 given our limited time and resources, but should be considered in the upcoming roadmap for 2010.

Grow our SUMO community

This has been an ongoing focus of the SUMO project in 2009 and will continue to be the case in 2010: a healthy and strong support community paves the way for a great web experience for our users (which is what the SUMO project is all about).

The specific focus in Q4 should be to fix bugs and add features to SUMO that has the strongest potential of having a positive impact on the community volunteers' experience on the platform.

Furthermore, we see opportunities of community growth in the upcoming Firefox Support for Mobile website; bugs related to this SUMO-powered support website should be taken into consideration as well.

Proposed SUMO Q4 Goals (DRAFT)

A. Upstream SUMO Patches to TikiWiki

See the proposed plan.

If needed, this goal could be further split up into:

A.1 Upstream major SUMO features to TikiWiki

This would include the features listed under Features / enhancements that should go to Tiki.

A.2 Upstream remaining SUMO patches to TikiWiki

This would include all other bug fixes/changes that would be needed in TikiWiki in order to successfully upgrade SUMO to Tiki 4.1/5.1 in the future.

By splitting the goal up into two sub-goals, we could give them different priorities and allow the goals below take priority over A.2 (or even move A.2 to Q1 2010).

B. Improvements to search

This had to be pushed out of SUMO 1.4 because of lack of resources, but the feature would greatly increase contributors' ability to find info on SUMO.

We have expanded our goal to entail upgrading the Sphinx search engine (AMO is already running the newer version) and a rewrite of the way we do indexing -- we can shed the XML generation and speed up indexing which will allow us to reindex more often.


  • Improved user experience with to more up-to-date searches
  • Improved community experience with advanced search form
  • Reduced machine load

B.1 Upgrade Sphinx

This will make the advanced search features easier to implement, and enable us to rewrite the indexer.

B.2 Implement Advanced Search

Finish and close out the remaining search related bugs from 1.4.

B.3 Rewrite indexer

We will rewrite the indexer to not use XML. This will speed indexing and allow us to index more closely to real time, improving user experience.

C. Maintenance Release to Improve Community Experience

The goal is to fix bugs and add features to SUMO that has the strongest positive impact on the community volunteers' experience on the platform today.

The SUMO team will work on identifying this list of bugs together with the community. Special consideration should be given to bugs/features reported by or confirmed to be frustrating/lacking and as such making it harder to use the support platform. Suggested areas to cover:

  • Remaining high-priority ForumUX bugs that were pushed out of 1.4.
  • Other long term bugs affecting the community negatively.

The scope of this should be carefully controlled. Based on experiences in Q1 and Q3, we should limit this release to approximately 30 bugs.

In order to further increase focus, this goal could also be split up into separate goals depending on the specific bugs to fix. Releases might then be as specific as:

C.1 Fix regressions and usability issues caused by SUMO 1.4

See SUMO 1.4.1 bug list

C.2 Fix remaining, essential ForumUX bugs

Specific bug list to be decided.

D. Rewrite one or two components of the TikiWiki core

Rewrite one or two components in TikiWiki that makes the platform unnecessarily hard to work with today. The result would be a win for both the SUMO and Tiki project.

Background: The current SUMO codebase is often frustrating for SUMO developers to work on. If we could identify one or two things that would improve the code and make the platform more fun to work on, that will make it easier to add new features that enhance the platform for users and our community in the future.

Potential areas to rewrite:

  • helpers for templates, helpers for controllers
  • cleaning up routing

Webdev team will decide which areas to focus on and formalize this goal to determine resource needs (although a reasonable estimation from morgamic is 2 weeks work for one person).

Support/Q42009 Goals/Goal D Candidates

Proposed order/priority

Based on the proposed goals above, here is a potential order/priority:

  1. C.1 Fix regressions and usability issues caused by SUMO 1.4 - SUMO 1.4.1
  2. A.1 Upstream major SUMO features to TikiWiki
  3. C.2 Fix remaining essential ForumUX bugs
  4. B.1 Upgrade Sphinx
  5. D. Rewrite one or two components of the TikiWiki core
  6. B.2 Implement Advanced search
  7. B.3 Rewrite indexer
  8. A.2 Upstream remaining SUMO patches to TikiWiki

Ideally, A1 and A2 would happen in parallel with the help from LPH of the Tiki community. If not, B3 and A2 should be considered "at risk" by default.