View source for User:Bhashem/AddOns-AMO-Principles

This short document outlines the top few principles around Add-Ons, the Add-Ons ecology and addons.mozilla.org (AMO). They bolded items below represent the most important goals of what AMO needs to be.

= I. Create the most comprehensive Firefox Add-ons directory on the web =
* Create nodes for non-AMO hosted stuff
* AMO search results should include non-AMO hosted stuff
* Write a crawler that collects xpi's and metadata and populates AMO. (e.g. Mahalo just launched something. Stuff is in private beta, perhaps you add it to the sandbox or a non-AMO portion of the site)
* Create a manual index of individual locations (Google, Yahoo,Skype,selected user blogs, etc...)
* Allow publishing from mozdev and other extension project development sites
* Suck in from other AMO-like sites
* Comprehensiveness include language and locale breadth, get add-ons and AMO translated into multiple languages

= II. Help various user types find and discover addons =
* Improved addons manager in Firefox
* Create a theme browser to quickly discover themes
* Improve search so that results and keyword yield good results
* Developer an AddOns Recommendation Wizard that is task based
* Allow for searching by Firefox architecture (plugin, toolbar, sidebar, etc…) – perhaps by clicking on an image of the Firefox areas?
* Recommendation Lists
** Enable publishing and sharing of recommendation lists from third-parties (RecoList Builder)
** Create recommendations based on browser data (e.g. search or browse history - we see you visited facebook, do you want the Facebook addon?)
** Create recommendations based on psychographics (e.g. love social networking ,photography, etc…)
** Get personal - currently all recommendations are not based on what you have already
* % of users who use this extension also found Addon X to be useful
* “You might also like” recommendations
* Allow users to self-identify (e.g. new user, advanced user, web developer, extension author) to provide recommendations

= III. Ensure the highest quality for what gets published =
* Safe from a integrity of software, user experience, security, performance, privacy, etc..
* In cases where AMO can't vouch for an addon, provide additional details/info so user can make an informed decision (e.g. this is why it's sandboxed, it's non-AMO hosted, etc...)
* Ensure users get a "safe" experience
* Security signature of extensions/SSL site
* Create a reliability rating system
* Enforce security reviews for extensions to prevent malware attacks (esp. upon previously approved extensions) – Not just new stuff coming in, but all updates
* Share a talos/performance/memory impact assessment (increase in startup time, page load and/or mem usage)
* Create tiers of review (e.g. Random AddOn, Basic Review, Security Checked, Privacy Checked, Spyware Free, etc...)

= IV. Create a satisfied ecosystem of extension developers, editors/reviewers, translators =
* Consider all the steps from learning to author an extension to writing it to deploy to supporting - dev tools, docs, tracking & stats, etc...
* Provide better stats/analytics tools
** Historical view graphs - # of users, growth rates, growth trends (bunch of time periods)
** RSS feeds that be subscribed to
** CSV that can be downloaded
** Funnel visualization (a la Google Analytics) - download, first run, active users
* Build value proposition of why it's good to host on AMO
** What you get for free? hosting, distributed worldwide, uptime, whitelist, a site with traffic, marketing?, real-time instrumentation, etc...
* For self-hosters, provide a set of tools for publishing success
** Here's an <object> or iframe that features your stuff integrated into your site - no AMO branding, etc...
** Here's a way to private label your stuff on the AMO site - e.g. upload you header/footer/etc... hosted on AMO
** For transalators, possibly work with Babelzilla
*** Didn't want to discriminate against other translation sites
*** Offer a developer API, write access - can access localization for addons
*** Add addon metadata translation to xpi themselves. For example, look for specially named file (might be a bug on it)

* [http://sethop.com sethop] suggests: Consider building community and participation around the review process. Much of Amazon's added value comes from the reviewer community, that would be a good place to look for interaction design hints.
** If reviews are rated useful/not-useful, then it becomes much more obvious what sort of review is useful/appropriate and reviews should improve across the board,
** Giving reviewers a track record / reputation metric based on their reviews will encourage quality reviewing, particularly in combination with good helptext about what constitutes a useful review.
** Consider lists of the most recent and most highly rated reviews and also the most active and most highly rated reviewers, with the ability to click through to see other reviews by the same user.  
** Also see [https://bugzilla.mozilla.org/show_bug.cgi?id=375649 #375649] and [https://bugzilla.mozilla.org/show_bug.cgi?id=384023 and #384023]

= V. Grow the site to include ALL types of customizations =
* Include all types of customization that can be done with Firefox:
** microformat detectors & handlers
** RSS feed and content handlers
** Plugins (harmonize with PFS)
** Personas
** Greasemonkey, CoScripter, webapp profiles
** Joey/Mobile/iFirefox home page Widgets?
* Customization technotes that don't involve installing new software