L10n/HomePage/Status/2007-08-15

From MozillaWiki
Jump to: navigation, search

« previous week | index | next week »

Meeting Details

  • 9:00am Pacific, 12:00pm Eastern
  • 650-903-0800 x91 Conf# 206 (US/INTL)
  • irc.mozilla.org #l10n-drivers for backchannel

Attendees

WilC, Choffman, AxelH, SethB, Rishi, TimR, Clint, Mic

Meeting notes

Minotaur update (30 min to 1 hour - Clint lead)

  • tool is in a decent ready state for L10n/Fx3
  • Clint will be moving off to Performance testing shortly so we need to use tool pretty much as is with minor adjustments
  • address specifications/new requirement post Fx3
  • Mic to review and approve the initial data dumps that minotaur extraction tool provides

New markets and data

  • Rishi been collecting good information and as we scale up India (and maybe Africa), it'd be cool to figure out what we should continue to collect and share (Rishi lead, Seth, Choffman add)
  • market focus has been 7 markets: canada, singapore, australia, mexico, pakistan
  • research has been focused on:
    • state of web (internet useage, top tech portals, top visited sites, etc)
    • state of community (history in area, who is there, etc.)
      • who (start on l10n home pages); Rishi asked how do to get in touch with them: on l10n home page; wiki pages with link to community page using armenian example, let them choose their tools but at least use wikimo as coordinating/information sharing mechanism)
      • user groups, (IRC, newsgroups, mailto lists); Rishi can also look at local wikipedia's for who is editing them to connect to open source community (likelihood is high that in each country these people will know each other)
    • what have they done e.g., localization, inside open source world, areas contributing to, adding code, other open source networks do they belong to
    • do they know other mozilla contributors who are not localizers
    • events taking place of importance in their region; Seth's approach for the way he set up visit to India is the way to go (ACTION: Seth send me note to document how he did this)
  • if there aren't people there related directly to Mozilla: is there existing leadership in place that is/relates to open source
    • who are open source activists in region or groups
    • then ask all same questions
  • then with all these information look for most leverage for what our strategy needs to be in that area
  • for the regions we're looking at we don't currently have any localizers, objective is recruiting new support

Community development and the new dbase Asa proposed, what do we need to provide here, is there a time line (Seth - 5 min)

  • Seth/Rishi lead, Seth lead after he leaves
  • more work for design and importing data, then hand off to IT/Web Dev
  • start with just MoCo employees to access

AMO/Lang packs update for all.html site (made progress here, 5 min, mic)

  • We have the ability to make bundles of extensions+firefox. It's unclear how automated that can be.
  • Language Packs on the public side will appear on /Firefox/all.html.
  • We need to adjust the criteria editors use to approve language packs. Criteria should require reviews for technical quality and accuracy. General positive reviews are encouraged but not required. (Accuracy is the hard part here)
  • 391705 is the AMO policy bug for criteria
  • all language packs be they in sandbox or not should be promoted with clear usage caveats indicating state of build (mic to check in with John on this)
  • next steps on all.html are: Mic to draft text for main page and linked-to pages that include bundle i.e., switcher+lang pack +possibly Fx download
  • amo review process an issue, difficult to get reviewers, Mic/Wil to sort out criteria for reviews and ways to get other community members to review
  • translation quality is important otherwise it too adversely impacts UE so it's important it is reviewed
  • who do we get to do the review: anyone we can
  • need to get a framework for accomplishing the evaluation: get smartware to publish the framework they use and then recruit and make an easy process to say here's criteria for eval, steps for eval (broken out into couple hour tasks) and then get volunteers
  • this does not need to be a 30 hour review, we don't have to gate release onto 32 hour review
  • we can plug into litmus (once we extract template from smartware reviews)
  • important to expose the review so users are aware of it; it is a low bar to qualify
  • set minimum bar; want a criteria to say this is a 5* or 3* build
  • Axel proposes a technical review based on %of translated string
  • so long as doesn't have technical
  • need to make it visible enough
  • ACTION: Axel will have it go through technical review; if there is any additional code in it then Axel will R- it and doesn't go past that stage
  • create localization review guide to describe different states: on all.html need reviewing; web page we create; star system to incorporate feedback into * star

ACTION: mic check in with John Lilly on approach; mic/wil to create criteria

Content handlers (made progress here, 5 min, mic)

  • mailto, rss, news feed, webcal
  • spec still to come as well as policy for choice in locale

Firefox 3 preparation

  • idea Mic had proposed about regular language triage so we know current status all the time (what tools might we look at, could we have a scrum type agenda item, etc)
  • wiki page
  • new issues discussed on call
  • use bugzilla - create subordinate for each locale
  • axel is working way towards having data of state of all localization accessible
    • compare locales, how many string are they missing, how much time take those in progress to catch up with en-US stage
    • which languages take 1 week or 3 weeks, which languages haven't done anything at all yet; ACtION build a tool for that
    • how engaged are they currently
  • get message for current state, semi-frozen features
  • first iteration there on list and then go from there
  • post to l10n google group

ACTION: axel/mic draft email