Translating Software Strings and Translating documentation

I remember seeing a post on LinkedIn from a Translation Management System developer stating that translating software strings and translating documentation are the same thing. 

This is not accurate.  Software strings are very different from those in documentation and need to be treated differently during the localization process and in a Translation Management System.

I have often been asked by software developers to run leverages from translated UI strings to documentation and online help.   The leverage from such an exercise is really small, to their surprise always in the single percentages.   UI strings are short, one to five words making up the maximum text.     Documentation or online help is much more verbose.   TM segmentation works on the basis of analyzing an entire string or segment.    The TM matches from UI to online-help or documentation is all at the very low fuzzy match level as a UI string will usually be less than 50% of a documentation string in length.

Because of the nature of UI strings being short an often shy on context, Translation Management Systems need to be able to support the addition of two extra elements not as important to documentation translation.   To support Software Localization Translation Management Systems need to be able to support the addition of Context information to source strings for translation and also the ability to support an initial extra leveraging stage whereby the String ID is used to give a guaranteed Translation Memory match.