After a software or web application
has been localized, it is usually tested before being released to the target
market. This step is critical to prevent usability, functional, and linguistic
problems down the road.
Why are localized products tested?
Because the localization process can introduce linguistic and functional
problems in the software and often reveals shortcomings in the software’s
design, especially if it hasn’t been properly internationalized.
Those problems can be caused either
by over-translation of some system variables that are invisible to the target
user and are not to be translated, or by modified functionality, which
sometimes must be implemented to meet local market expectations or
requirements. Letter wizards and spell checkers are typical examples of
this.
Linguistic Testing
Linguistic testing confirms that the
text in the user interface appears correctly and completely and is not
truncated, mistranslated, or misapplied. This testing generally requires a high
proficiency in the target language and strong familiarity with the product
itself.
Some
problems that linguistic testing reveals are:
- Missing translated text
- Text that remains in the source language
- Mistranslations based on the context in the user
interface
- Translations that are cut off or wrap to the next line
- Inconsistent appearance or usage
Functional Testing
In functional testing, testers
follow prescribed test scripts to run through all aspects of the product to
make sure it functions as designed in the localized environment. Often
functional testing includes verification that the localized product is compatible
with various operating systems and third-party products. Functional testing
generally requires good knowledge of the target language and total familiarity
with the product.
Some
problems that functional testing reveals are:
- Incorrect error messages
- Localization bugs that may cause software crashes
- Incorrect or missing UI elements, graphics, and windows
- Errors generated by installing the localized software
on a localized OS
Requirements for Proper Testing
Localization testing requires both
source and target language versions of the product installed in the environment
that a typical user would use. This means testers must have the correct version
of the operating system, language, regional settings, required third-party
software, and more. It is critical for the source and target software builds to
match, so the expected functionality can be guaranteed.
Testers typically follow strict test
scripts that clearly define steps and expected results. Any bugs found are well
documented in bug-tracking software. When the bugs are corrected, the software
is re-tested to make sure the defect is removed.
In the past, localization testing
was largely a manual process, but many tasks have now been replaced by
automated software that can discover truncations and other defects. However,
linguists are still required to discover errors in meaning, context, and
function.
No comments:
Post a Comment