Siebel Global Deployment Guide > Localizing Global Deployments >
About Testing Globalized Software
This topic can help you to test the efforts of your software internationalization and localization teams.
To test the configuration, you first need to establish what your key platforms are for your database servers, your Siebel Servers, your clients, and your operating systems, for each language that you intend to deploy within your network.
You must test each supported platform, keeping in mind that there might be subtle differences or additional third-party software required in some cases.
Testing for Internationalization
Test your globally deployed applications to verify that all of the internationalization-related features function correctly.
For example, to test for character integrity during saving and retrieving data, you might create test cases that use a wide range of characters from your target languages. For a particular target language or set of languages, enter such data into fields and save the record. Check that you can search for the records and find them. Select the records and view them, checking that all of the data values that you input are unchanged. After entering, saving, and retrieving the data records in the target language, no square boxes or question marks should appear in the data.
Some of the issues to consider when creating test cases include:
- Test for character integrity during saving and retrieving data
- Test for user interface character integrity
- Verify interaction with native language operating systems and application response to locale settings
- Verify functionality with date format (including date-time fields) and time format
- Verify functionality with time zone support
- Verify functionality with number format
- Verify functionality with buffer sizes, check for overflow
- Test file names with international characters
- Verify accelerator keys
- Verify support for the euro currency and its symbol (€)
- Verify decimal symbol calculations and arithmetic operations
- Verify calendar functionality
- Verify list separator
- Verify reporting and printing operations
- Verify sorting
- Verify query or find operations
- Verify phone number formats
- Verify layout for name fields, for example support for middle initial
- Verify layout for address fields, for example support for state or province
- Verify postal code is not a required field for a country with no postal codes
- Verify measurement units, metric or imperial
- Verify interaction with third-party software
- Verify import or export operations
- Verify taxation system support
- Verify translations are not truncated or otherwise corrupt
- Verify that only translatable items are translated
Testing for Linguistic Quality
Use people in your field offices whenever possible to help you check the localization for linguistic accuracy and consistency. Test any custom terms with the field, and pick terms that will be used often. For example, if you use a term like currency rollup, then make sure that it is correctly translated for your field users.
Using Automated Test Software
Third-party software is available to automate some test functions, although it does not replace the need for human testing.
Automated test software is very useful to test a wide range of characters, because an automated test can check all of the possible characters in Unicode, for instance. If an automated test is combined with a random test-data generator, then it is also likely to vary the data values much more widely than a human tester will.
On the other hand, an automated test cannot easily detect characters that have been clipped or otherwise hidden on the user interface, where the human eye can detect anomalies very quickly.
Of course, it is important to use automated test software that supports Unicode data values and error messages containing failing data values. If several languages are being tested, then it will be most convenient if the test software can switch between each language automatically.
Build an efficient process for defect fixing between reviewers, translators, and the development team.
Make sure that your defect tracking system also supports Unicode data. It is very hard to explain a problem with data values in another language if the defect system does not allow you to put the failing characters into the title or the record of the defect report.