20 October 2005


getting up to speed

it's ironic. after working on a product for people writing software while I was at that big company for the past four and a half years, I actually had very little exposure to code. after all, our process was to convince developers to squeeze everything that needed translating into as few files as possible, and for those files to basically be collections of translatable text.
what's more, the tools we used for our work sanitized the experience. all I saw was what needed to be translated (although I know this wasn't always true for some of my colleagues).
now, I am looking at code all day. I'm not translating UI, either. although I worked on a small manual for one component on Monday, since then I have had lots of code. makefiles. CPP files. H files.
and I'm translating comment files. stuff we usually ignored at my old company.
of course, it makes sense to translate this stuff. the code was written by developers, and it's going to be read by developers who need to make more code based on it. so they need to know how the code they read was written.
but developers are rarely good writers. of course, they suffer from the chronic afflictions of spelling and grammar mistakes (as I do). but the medium in which they write demands that they be terse. when the lines end, they can't just keep typing. they either have to finish a sentence (or a thought, even) within less than 80 spaces, or pretend they're on a typewriter: hit the carriage return, the space bar a few times, and then start typing again.
so they write in short bursts. calls function. returns value. very stacatto.
and if they happen to be verbose, which can be good for the reader, it's not good for the translator. although short sentences can be hard to work on (where's the verb? which one is the subject? what tense are we in?), long sentences have two major problems, one of which is brand new to me.
the old issue is the increased complexity. Japanese syntax is, in essence, the reverse of English. yes, in a simple sentence, the subject usually comes first (but readers rely more on little particles, parts of speech that mark the function of nouns in sentence). the complexity comes when clauses modifying nouns or phrases move in. those clause precede the modified part, unlike English, where a modifying clause usually comes after (with a which or a that marking it). so, in Japanese, you can be reading along and suddenly realize that what you just thought was the main bits of the sentence, is actually a modifying clause, 'cause here come the main buts, but wait, this is a modifying clause too...
parsing the sentence is half the fun, I suppose... :)
but when a long sentences has these line breaks in the middle, as code forces a writer to include, the software that we use to track our translations is thrown a curve. the way translation memory software works is it breaks down the original text into units. usually a unit ends with a period. or a colon. or a line break... uh-oh.
if Japanese syntax is the reverse of English, what this can mean is that I have to translate the part of the sentence at the end, sometimes three or four lines down, first. the translation units don't really match, then. so we can't really consider long sentences an effective part of the translation memory. which means the translator relies on the older version of translation memory: gray matter. :)

some other things to report.

my new work place has a nice cafeteria, though I've only gotten a drink there (Crystal Light! hurray!). I should probably reserve comment until I actually eat something. I bring my lunch. Monday it was microwaveable spinach curry. Tuesday, chicken soup. Wednesday, leftover oden (a Japanese dish where you boil foods together in a clear soup base of fish and seaweed stock. I made mine with tofu and fish cake, which I love). I also bring my Crystal Light packets and my protein bars.

there's a store at the new company, and they sell refurbished equipment, cheaply! Hiro is excited about that, as I knew he would be.

finally found my iPod so I can listen to my music at work, instead of webcasts from radio stations, which are nice, but the quality.... headphones are a necessity. after working in an office with a door for so long, it's hard to concentrate when there are other people around. I used to do it in Japan, though, so I'll probably eventually wean myself from the iPod.

music note: next year is the 250th anniversary of the birth of Mozart. Wolfie fans, unite! I have the complete Mozart recordings Phillips released several years ago at the 200th anniversary of his death, so I need to get those onto my iPod for the celebrations! I nominated three of his pieces for a Classical Top 40 at a local radio station, but one of them is not well known. it's called A Musical Joke in English, and it's Mozart's parody of the shlock composers of his day. if you've ever taken a class in composition, it's really funny to listen to, because it makes all the mistakes they tell you not to in class. and it sounds so awful in places, I just laugh out loud. I know, not a well-known piece, or one people might like upon hearing, even, but it's always been one of my favorites...

okay, time to go to work!

Comments: Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?