L10n/HomePage/Status/2007-12-19
« 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
Chofmann, SethB, TimR, Axel, Mic, Wil, Pascal
Draft Notes
- QA: Smartware testing
- some teams not happy with output from smartware (last year e.g., Italians didn't appreciate results); french team checked website of smartware to see what qualifications they have to provide useful output (makes it easier to dismiss as they are not good); if we don't think smartware is good enough then might be hard to find another
- possibly get feedback, do localizers think they can do a good job?
- Tim's proposal: go through locales we did last year and do a quick assessment of did we add value or not; do we still do these with smartware, find another company, etc.
- Smartware treat as another community output and allow them to file bugs directly
- Could start this week
- Small budget for augmenting community testing, trying to figure out best way to allocate funds we did research to find smartware to provide this service and it seemed to be high quality last year. we'd like feedback from all team smartware tested on. this year we will apply smartware to ; we will always use localizer assessment over a 3rd party
ACTION: Tim enable start with French, Polish and German and get feedback from localizers on usefulness; Mic post to Google group with feedback from l10n-drivers (links to reports produced last year and bugs)
- L10n-late tag for triage
- Benefits:
- better idea of when can start with QA testing
- which areas are locked down
- would also give us metrics over time
- mconnor feedback: late-L10n wrong key word to do this; currently no way to do another pass of bug list
- how do we tell localizers what's going on and when to do things vs last minute; best people to tag are engineers who are doing regular triage
- Benefits:
ACTION: chofmann sent mail to schrep on this topic; schrep will apply pressure to mconnor
- Fx3 dashboard
- how much work left to be done: % strings translated (Axel's dashboard mock up does this all right)
- how fast are localizers responding to changes required: hard to measure
- number of users in locale
- web services review
- Web parts and google snippets review
- litmus and community tests
- Number of localizers
here's the place where collection is happening axel doesn't want to do stuff manually two sets quantitative that we can reach through automation qualitative: manual automated systems can produce output but someone still needs to review it someone has to look at results of automation ACTION: Mic email here's list of stuff you want with priorities and sources of data and one dashboard
- https://bugzilla.mozilla.org/show_bug.cgi?id=405364
- place to put public list of sites incompatible with Firefox
- can report on mozilla.org; user interface localized; want to use the reporter tool
- don't want to change reporter just interface
- Wil work with <name>
- Tools with Dwayne
- spoke with Dwayne this am; in our opinion he's been very collaborative and helpful
- grant does not require matching (or any other type of challenge grant); has a significant grant to help pan African network exchange; research director to help allocate and spread funds; goal to localize software into more languages
- Next steps: Dwayne will provide list of languages this might impact (likely at least 12), list of other software or organizations that will use tools he's developing, and any revisions to funding request; Mic/Seth to provide a timeline for when we do what
- Candidates/Fx 3
- targets for new localizations Firefox 3; even though Fx3 is not stable given how long it takes to ship new language on Firefox 2 more confident that this is good route
- Chofmann/Mic good idea; Chofmann suggests also add why you should move to Firefox 3 vs 2 (cause most of our resources over Feb/March 2008 will be focused on Firefox3 and it take long time to get through process so best to focus al that effort on Firefox 3)
ACTION: Axel put together list of language that have a shot to make end of January or mid January/Fx3 time lines then Mic/Pascal will do their work to enable release; sprint for languages that are close if you're not close then go for Firefox 3