Instructions

Below you'll find the Open Session/Innovation Fair proposals that the L10n Drivers have identified as sessions we'd like to give at this year's Mozilla Summit. If you have a session proposal that you'd like to make and do not see here, please add it according to the format of the other proposals. On Monday, September 16, we'll discuss these proposals, determine which locations they should be presented in and by whom, and then will submit them for approval and scheduling.

Localizing with L20n

Your name

Zibi "Gandalf" Braniecki (Toronto), Matjaz (MV), Stas (Brussels)

Summit location of session

All

Short description (1 paragraph)

As we're transitioning our products toward new localization framework, this session will help those interested in localization get on board with the new technology and brainstorm its future.

Anticipated outcomes

• Understanding what L20n is
• Understanding how to work with L20n
• Help build a feature set for the next version of L20n

Outline etherpad

Summary for L20n

Language choice on Mobile

Your name

Jeff (Toronto), Axel (MV)(reporting), flod (Brussels)

Summit location of session

All

Short description (1 paragraph)

Mobile devices come with constraints unknown to desktop machines. As such, localization contains different requirements in order to successfully produce a localized mobile experience. We'd like to discuss these constraints and strategies the L10n Drivers and the localization community would like to see in order to win in this new ecosystem. Additionally, the entry point for localization teams in the Mozilla localization ecosystem has been localizing Firefox desktop. As the world changes toward increased use and availability of mobile devices, it's time to discuss adding mobile Mozilla products as entry points to localization for relevant locales. We anticipate this to not only change the localization landscape on mobile, but on Firefox desktop as well. We'll discuss here the localization communities interest in this approach and how this might impact Firefox desktop localization for smaller locales.

Anticipated outcomes

• Increase awareness of technical l10n issues for all Mozilla mobile products.
• Get feedback from the community on shipping Mozilla mobile products as an entry point for new locales.
• Increase awareness of l10n shipping issues for all Mozilla mobile products.

Outline etherpad

Summary for Mobile

This ended up passing down information to localizers in most locations. The session in Santa Clara had Amir, who brought in a bunch of data for the language selector as well as fonts.

Next Actions:

• Drive bug 881510 (Accept-Language header customization UI for Android) into the tree. Ian Barlow is having mock-ups in the bug by now.
• Check how much of that UI is useful for the UI language choice on Android.
• Propose a port of that UI to Firefox OS for language choice in general (there's no UI language on FxOS)

Localizing Firefox Desktop

Your name

Jeff (Toronto), Axel (MV), flod (Brussels)

Summit location of session

All

Short description (1 paragraph)

What does Firefox desktop consist of as a product? Is it devtools, error consoles, UI, support, or all of the above? Are the features customizable by region and who makes those decisions? As localization of Firefox desktop evolves, the L10n Drivers seek to understand the community's definition of Firefox desktop as well as identify and discuss how changes to the Firefox release schedule impact localization and what technical and linguistic elements are expected in order to ship a localized Firefox desktop product.

Anticipated outcomes

• Better define what localizers expect when localizing Firefox desktop.
• Get community feedback on new approaches to localizing Firefox desktop.
• Get community feedback on new approaches to shipping language packs.

Outline etherpad

Summary for Desktop

L10n quality on desktop

• The Grandma rule - If I can't explain it to my grandma, the translation's too complicated.
• Aiming for a mixed register on a per module basis.
• Excitement about upcoming style guide, but concern that it will be too long.
• Low quality localization = low reputation/quality of Firefox
• Localizers generally do not update terminology and term consistancy because they don't know how and it's too hard.
• Localizers would like test cases for testing l10n.

Who are we localizing for, ourselves or our users?

• Desire to understand users, but difficult to let users know who to contact to give that feedback.
• Unaware of how to get feedback from users. Possible to use input.mozilla.org, or a version of this?
• China seems to be able to do this well.

Langpacks as localized products

• Langpack install, selection, and update experience needs to be more comfortable for users (i.e., breaks the Grandma rule).

Defining Firefox product

• In favor of splitting up Firefox into translatable/non-translatable buckets.
• devtools and dom are unnecessary.
• Language fallback issue divided; many in favor of English.
  • Good point for English: you receive more Google search results for confusing error messages in English.

Contributing to a localized Mozilla

Your name

Chris (Toronto), Pascal (Brussels), Matjaz & Arky (MV)

Summit location of session

All

Short description (1 paragraph)

Since the last Mozilla Summit, the Mozilla localization community has grown extensively. Contributing to the localization effort has attempted to evolve with the community's growth, however, many elements of the l10n contribution path remain a mystery and unresolved. We'll discuss which parts of contributing to Mozilla l10n are challenging, alienating, and require selective skills. We'll discuss tools, strategies, and resources for building localization teams. Finally, we'll seek to come to a concensus on the expectations of all participants within Mozilla localization, from the L10n drivers to regional l10n teams to individual localizers.

Anticipated outcomes

• Teach l10n teams techniques and strategies for finding new localizers and keeping them engaged.
• Allow l10n teams a platform to share their own growth experiences.
• Discuss the roles of the L10n Drivers, l10n teams, and any changes that need to be made in role expectations.
• Discuss what elements of the l10n process are inhibiting l10n community growth and strategies to overcome them.

Outline etherpad

Summary for Localized Mozilla

Community building

• Some locales confess that the reason their l10n team isn't growing is because they're not proactively building it.
• A lot of people had never thought of actually contacting localizers from other projects for their locale to look for help.
• New localizers are missing information on best practices/mentorship/guidance. For example, Goran is a new Slovenian localizer and still learns, so when dealing with web dashboard he mistakenly downloaded and translated the whole English source file, while only 5% of the strings would have to be translated if he downloaded the Slovenian file.

Pros & Cons

• A few comments about too many tools and that elmo dashboards are difficult to understand.
• A couple of localizers confirmed that teaching l10n processes to people that don't know much about open source tools in general is hard.
• As much as I tried to focus discussion elsewhere, everybody kept saying that the l10n process is too complicated to grow communities, so a lot of Santa Clara discussion was about tools.
• Copying strings from files to files when localizing Firefox is unacceptable for the Indian community (I even captured the word insulting).

Tools to improve localization quality

Your name

Axel (MV), Pascal (Brussels), chofmann (Toronto)

Summit location of session

All

Short description (1 paragraph)

Many localizers have noticed a technical gap in the localization tools and resources offered to perform localization at Mozilla, specifically in assessing and maintaining translation quality. Some regional communities have attempted to organize and create these resources, like the FUEL effort in India. The L10n drivers have dedicated time to evaluating tools that incorporate translation memory features, more precise metrics, as well as efforts to create and use glossaries, termbases, and style guides. As a community, we'd like to discuss requirements for such tools and resources from all participants in the localization effort at Mozilla.

Anticipated outcomes

• Provide a platform for l10n teams to discuss their own translation quality control measures.
• Propose methods and tools for improving translation quality and get feedback from the l10n community.
• Identify needs in l10n quality tools that scale to the community.

Outline etherpad

Summary for Tools

The Role of Commercial L10n at Mozilla

Your name

Jeff (Toronto), Pei (MV), Pascal (Brussels)

Summit location of session

All

Short description

• In engineering?
• In marketing?
• As a volunteer partner or paid vendor?
• L10n drivers & community as a internal service providers in competition with paid vendors
• How to improve commercial l10n work through language resources and centralization.

Anticipated outcomes

• Building a consensus with localizers on how to use Mozilla resources on localization.
• Discussion on the role of commercial participation within the l10n process.
• Clarify the role of a commercial Mozilla localizer.

Outline etherpad

Summary for Commercial L10n

Current problems with some solutions

• Absence of language resources to provide to vendors. Garbage in, garbage out. Guidance on creating and maintaining these resources is desired.
• Localizers want to be involved earlier in the planning process to keep Mozilla reaching out to vendors. Let the community decide what can/should be localized by vendors.
• Partners are making final calls. This should be transitioned to the community.
• Vendors are at a larger contextual disadvantage than the community. Requiring vendors to work/consult with the community would resolve this.
• Other areas of Mozilla contract vendor work. How do they do it? What can we learn from them an apply?
• Community should have final approval on vendor work.
• "getting web agencies to step in won't solve the lack of time to organize a marketing campaign" - Alizée
• stas mentionned the idea that we should accept that campaigns don't get localization and learn from that and wonder why some campaigns don't generate interest in our core community instead of making sure that we localize everything
• Money spent on paid translations should instead be funneled to building l10n teams.

Community's perception of working with vendors

• Uncomfortable with using vendors to run QA on l10n work.
• Most felt comfortable with vendors collaborating with community to create language resources.
• All hate the idea of a vendor being used to maintain a locale when it falls behind.
  • "Could lead to the downfall of the community infrastructure as we know (and love) it."
  • IT guy said that introducing paid ressources means we will pay forever, every time IT paid for a ressource to complement their community, they killed their community in the process
• Liked the idea of vendors becoming part of the community.
  • Everyone is then on equal ground.
  • Allows existing community to find new contributors with training and skills.
  • Opens the potential for existing community to teach vendors about openness on the web.
  • Opens the potential for existing community to receive contracts from vendors for translation projects.