> [: Daiki Ueno :] > [...] as well as that the same tree structure is preserved after > translation. I wonder if the latter check might be too rigid, but > sometimes it would be useful.
It is too rigid. Intentional changes to markup structure in translation do happen in practice. One simple case is when the translator is both well-versed in the given markup and more picky about semantics than the author. One may claim that this is wrong way to do it, and that the translator should instead take it up with the author, to improve the original first. But in reality the communication budget of the translator is limited, and gets expended on more pressing matters. The more serious case is when the original markup structure may be fine for the source language, but would work poorly in the target language. An example: msgid "" "From the context menu you can choose to <guilabel>Remove</guilabel> or " "<guilabel>Deactivate</guilabel> the selected frobaz." In languages with more sophisticated verb declension this would likely not work, and translator could produce the equivalent of: msgstr "" "The context menu allows you to remove or " "deactivate the selected frobaz." with verbs having proper declension for the sentence, different to those seen in the context menu. -- Chusslove Illich (Часлав Илић)
signature.asc
Description: This is a digitally signed message part.
