TM-based localization - Mozilla experience

Revision as of 07:53, 7 November 2011 by Smo (talk | contribs) (Created page with "Translation Memories, or TMs are now a given. Translators working with TMs report they work, on average, between 20-30% faster than without TMs - so far Mozilla l10n was spared t...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Translation Memories, or TMs are now a given. Translators working with TMs report they work, on average, between 20-30% faster than without TMs - so far Mozilla l10n was spared this expertience (sg). I want to show where these 20-30% can be found when it comes to Mozilla localization. Here's an incomplete list of use cases:

  • pretranslation: you probably had a deja-vu experience of "I translated that several times already..."
  • quality assurance: your team does not consistently translate "folder" and "directory". How do you make sure the distinction is kept in the translated material, whoever has done the translation?
  • fuzzy search: your present system suggests a translation - but you had a better one some time ago, that would better fit the context...
  • support for a large number of file formats: what about a tool that allows you to localize in all kinds of different format (PO, but DTD, properties, html) while keeping all translations in one single reusable file, you can share with other members of your team?
  • maintaining your translations - how long does it take you now to get a minor change in the source file percolate into the staging area?