如果没有为可译性做早期设计就在事后进行翻译/本地化，那么我们只有对所有的东西进行彻底的再测试。(一致同意) 这时，基线语言自动化脚本就体现出很重要的价值。
33. We disagreed with the following statement: It is important to regress all bugs in a localized version and, to the extent done in the baseline version, to extend automated tests to establish the same baseline for each language. (7 no, 1 yes.)
我们不赞成下面的说法：按照基线版本中的样子回归本地化版本中的所有bug，以及通过扩展自动化测试为每种语言建立相同的基线版本都很重要。(7票反对，1票赞成)
34. The kinds of bugs likely to arise during a well-planned localization are unlikely to be detected by baseline regression tests. (9 yes, 0 no).
在规划良好的本地化过程中，可能会增加一些不容易被基线回归测试探测到的bug。(9票赞成，0票反对)
35. We didn’t vote on these points, but I’ll insert them here because they provoked thought and discussion (and because I think Marick’s on the right track.) Brian Marick suggested that in planning for automated localization testing, we should be thinking in terms of several distinguishable classes of tests. Here are some examples:
我们没有对下面的观点投票，但是我还是在这里列出它们，因为它们能引起思考和讨论(还因为我认为Marick的想法在正轨上)。Brian Marick建议在规划自动化本地化测试时，应该考虑一些可区别种类的测试。下面是一些例子：
· Language-independent automated tests, such as (in many but not all cases) printer configuration tests, other configuration / compatibility tests, and tests of compatibility with varying paper sizes.
语言独立的自动化测试，例如(在很多但不是所有情况下)，打印机配置测试，其它配置/兼容性测试，以及不同纸张大小的兼容性测试。
· Specific-language automated tests, if these are worthwhile. If you expect to keep selling new releases of your product in French, German, and Spanish versions, you might find value in creating some French-specific, German-specific, and Spanish-specific automated tests.
如果觉得值得的话，可以使用特定语言的自动化测试。如果你希望继续出售新产品的法文版、德文版以及西班牙文版，你会发现创建法文、德文以及西班牙文自动化测试是很有价值的。
· Most tests that are language specific will probably best be handled by manual localization testing.
特定语言的大多数测试有可能被手动本地化测试处理好。
· International-specific tests that are handled by automation. These tests check the translatability and localizability of the software. For example, you might provide dummy translations that use strings that are too long, too short, etc. You might provide text that will be hyphenated differently in different countries.
特殊国际化测试能够被自动化。这些测试将检查软件的译本和本地化情况。例如，你可以用很长或很短的字符串提供虚假译文。你也可以提供不同国家中用不同连字符连接的文本。
· Cheaply localizable tests. Marick’s expectation is that this class is small. However, some tests aren’t concerned with the strings used or the screens displayed. For example, stress tests to find memory leaks depend on repeated executions of some functions, but the text and graphics on display might not matter. Localization of some of these tests, to the minimum degree necessary to make them useful, will be easy.
简易本地化测试。Marick希望这是一类小型测试。但是，一些测试并不与字符串或屏幕显示有关。例如，强力测试能否找出内存泄露取决于一些函数的反复执行，而这与显示文本和图形没有关系。把这类测试按照使它们可用的最低程度本地化来将会很容易。
The bottom line is that even if you have an extensive suite of automated tests for an English language product, this might not speed you very much when testing a translation.
本地化的底线是，即使你为一个英文版产品创建一个自动化测试扩展套件，它也能够让你非常迅速地测试一个译本。
文章来源于领测软件测试网 https://www.ltesting.net/
