Jetpack/Features/2012-02-24

From MozillaWiki
Jump to: navigation, search

Details

  • Time: Friday, 2012 February 24, 9:05 - 9:50am PT (17:05 - 17:50 UTC)
  • Location:
    • Audio/Video: Jetpack vroom
    • Audio Only:
      • US: +1-800-707-2533, password 369, conference number 99449
      • US/Intl: +1-650-903-0800 or +1-650-215-1282, extension 92, conference number 99449

Agenda

  • actions from previous meeting
    • dietrich to discuss adw's "jetpack in core" proposal in jetpack discussion forum
    • dietrich to talk to gavin and mossop next week about timeline for "jetpack in core"
  • Apps feature in Firefox 13
  • pdf.js status

Attendees

  • dietrich
  • myk
  • gozala
  • krizsa
  • dcm
  • kwierso
  • ejpbruel
  • ochameau
  • canuckistani

Minutes

actions from previous meeting

dietrich to discuss adw's "jetpack in core" proposal in jetpack discussion forum

  • dietrich posted it
  • not a ton of discussion
  • people are sanguine
  • dietrich talked to irakli
  • he had suggestions for how to implement loader
  • next step sitting down w/irakli and developing concrete set of steps to make it happen
  • -> dietrich and irakli to come up with plan for landing jetpack in firefox


dietrich to talk to gavin and mossop next week about timeline for "jetpack in core"

  • didn't happen; we'll wait until we have a plan

apps feature in firefox 13

  • some aspects of feature don't seem like it makes sense to implement as addon (f.e. launcher in home tab)
  • feature set is not cohesive unit
  • seems like it'll be easier and make more sense to land in core
  • jetpack advantages seem more around development workflow and architecture
  • folks miss apis for deep integration
  • but people could write their own apis, there just isn't a good place to share them
  • jetpack team is talking about sharing modules this week, hope to have plan soon

making changes once jetpack is in core

  • gozala wants to make sdk packageless, that'll be a significant change
  • if it'd be hard to do after landing the code in core, it should happen first
  • -> irakli to start discussion about packageless sdk proposal

pdf.js

  • currently being integrated as addon
  • has some localization issues around build system tooling
  • other issues include performance, accessibility, and memory consumption
  • so far no known issues due to it being an addon

addon-based feature development principles

  • people should know why we're working in this group
  • -> myk & dietrich to draft mission statement/statement of principles (in four weeks!)

landing in core

  • jetpack developers will continue to develop the code
  • jetpack reviewers will continue to review code changes
  • once landed, changes to apis necessitate updating core code that uses them
  • the manifest is for security, but that doesn't necessarily make sense for core features
  • -> irakli to propose manifest strategy for core features using the sdk

hiding internal functionality

  • ways to get people not to use internal functionality that we want to be able to break
    • don't expose it as an api
    • put big warning signs on it
    • prefix its symbols with underscores

superreview

  • core apis typically need superreview for api changes
  • it's important to define what we mean by superreview, as its meaning has changed over the years

Actions

  • dietrich and irakli to come up with plan for landing jetpack in firefox
  • irakli to start discussion about packageless sdk proposal
  • myk & dietrich to draft mission statement/statement of principles (in four weeks!)
  • irakli to propose manifest strategy for core features using the sdk