I wrote earlier about how I thought when translating a few lines in Shinimasu. This series is going to be in that vein, with an eye to explaining translation decisions and highlighting unusual takes. I’m going to try to make it interesting for people not knowing Japanese, but to save effort I’m not going to be providing literal translation equivalents to lines.
Why am I doing this? Because my brain is a fuck and producing blog posts is an interesting motivation for doing a second pass on my translation.
Unfortunately for those expecting worthwhile content I feel like digressing a bit into history and methods for this first post, though. This is what my TL setup has looked like for most of the time I’ve worked on the project:
I started out doing 64 lines in December 2017, this got Asonn involved, and he introduced me to Porygon, who set up a git repository* and provided the tool you see. My brain swears I tweeted this pastebin, and I know I at least got some comment, but twitter search can’t find it so who the fuck knows? Anyway, I probably did 129 lines just copying from the game or script (can’t remember), then I copied them to the tool and worked there. One of the joys of working with porygon is that he has highly motivating auto-updating progress pages for you to fap to after pushing your new lines. This probably helped me more than I’d really like to admit.
Either way, apart from being convenient for reinsertion later**, the tool has rudimentary edict-lookup of the (autoparsed) tl lines, which is convenient if you’re extremely fucking lazy. I’m not going to say I never used it (I am extremely fucking lazy), but going j-j definitely was needed more than once. Other than that I guess it’s ok, though it does have a still-unfixed bug where it’ll fuck up and display too few lines of text in a box due to some miscalculation.
It’s certainly missing some features my dream tool would have, though. Personally I’d love to be able to see the script commands surrounding a line through some UI element to expand, as this could partially substitute for actually having the VN open for visual/scenographic context. It doesn’t have EPWING lookup, but that’s high effort since the format is bullshit apparently. It also doesn’t let you play voiced lines associated with spoken lines, though Shinimasu is unvoiced so I guess it doesn’t really matter for this project.
Today I had to contact pory since it had stoped working properly; it turned out my build of the tool was old enough that a bug with java 9 (I had recently updated) was making it unusable. He quickly got a fix for the tool, but it took enough time that I lost the energy for revising my tl. Or that’s my excuse, anyway.
See you next time for actual tl discussion w
*What’s a git repository? Well the long answer is long and full of programmer-speak, but basically it lets you keep an online backup of your files, preserving older versions each time you decide to add a newer version to the server. You can do this while multiple people are working on the same file sometimes, though it can get hairy. I ended up not needing this much, but it’s been good insurance against data loss (and I have changed laptops at least once during translation, also had to reinstall windows once…). Really if you don’t have a backup for any translation of length, you’re probably doing it wrong (but also I am a CS student so it’s… not as hard for me w)
**By virtue of saving the line number in the original script where the Japanese line was and associating that with the eventual translated line. I used a simplified version of this myself based on google sheets columns when I did tech for the ichigo & kyuugo tl.