Translation is as much art as science. It remains so, and probably always will, even after all the advances in linguistics. Even as machine translators continue to improve, so far none have managed to best the work of a human being well-versed in the given language and the culture that originated it. Even between Western languages, primarily Latin-based with a little Ancient Greek, there exist certain ambiguities and peculiarities that bedevil even the most sophisticated programs. For instance, German possesses many words that simply do not have an English equivalent. Weltanschauung may directly translate to 'worldview', it is more accurate phrase it as 'comprehensive worldview'. But to convey whole meaning of this single word requires a whole sentence; 'the fundamental position and viewpoint of an individual, group, or society, comprising the entirety of the individual or society's objective knowledge; beliefs; feelings, and values-as well as the framework through an individual, group, or society interprets and interacts with the world.'
It becomes apparent that a translator's job is much more difficult than simply substituting words. If it were that simple, machines would have replaced translators decades ago. Aside from grappling with syntax, grammar, etc, the thing that separates a good translator from a mediocre one is their skill in conveying meaning. A great translator must also be an amateur historian; sociologist; psychologist; anthropologist; and a skilled writer themselves. Parsing out the connotations, not merely the denotations, of words. Computer programs can aid, but never replace, a human translator.
If German poses such a formidable challenge, one must stand in awe of the challenges of translating from Japanese to English-from any Eastern language to any Western language. Here, the ratio between art and science almost entirely favors the 'art' component. It's not merely different kinds of words; it's a different sort of language all together. A machine translation from, say, French to English may be clumsy, but they at least share a common origin. So, a machine translation is at least readable. Translating between East and West (on either side) is a long and involved process. Here, it's more important to ensure meaning isn't lost in translation, as opposed to exact wording. As a result, multiple translations are possible.
I don't know any Japanese beyond the basics. I'm not a linguist or a professional translator, but one doesn't need to in order to know if a translation can be better.
To cut to the chase, I don't think Rogue's translation of the MGQ trilogy is good enough. I am not insulting his skills; he's no amateur. It's not a machine translation; but it clearly falls short of the state ideal. I don't fault him for it either; he's a busy guy, and it's great work in light of that. It is for that reason that it fall short: a lot of the dialog feels stilted, lifeless, not suited for a given character's personality. A fan community like ours should be buzzing with countless translation patches, each with their own strengths and weaknesses, groups and individuals sharing and critiquing each one's work. In addition, it would make Rogue's work that much easier; the more work he can draw from, the better his translation can be.
The tools for unpacking the various files and scripts are easy enough to find: just google NScript or ONScript. In the future, I'll put up a section of my profile, linking to resources of interest to any translator who feels the call of duty. Pre-patch, we have access to the original Japanese text. One doesn't even need to reinvent the wheel. For most, it could simply be a matter of improving on Rogue's patch, using the original text as an aud. It would also give us something to do until Chapter 3 is out. So, this entry-my first entry-goes out to all serious translators in the fandom. May you heed the call.