I hate this trend of everything needing to be on fucking #Discord
Fuck your little private treehouse club that’s badly managed and barren because you learned the hard way community management is a full-time job too much to handle for a single person
Use better, open tools ffs
Dungeons and Dragons is a role-playing game that lets you live out such fantasies as:
• Having money
• Making close friends as an adult
• Travelling the world without crippling debt
• Being able to change the world
• Getting better at something with practice
• Getting 8 hours of sleep each night
Things at #Bandcamp could clearly be better here. (And as always, support @bandcampunited ) https://aftermath.site/bandcamp-just-trying-to-keep-the-lights-on-following-epic-sale-layoffs
As a reminder, translation is a lossy process.
This post focuses on Japanese anime and manga, but I think the ideas are still fairly general and would apply to other translation work.
The idea that you can translate something word for word and “preserve the original meaning” is fundamentally flawed. It leads to broken grammar, awkward word choices, and poorly-flowing scripts. Ironically, this style of translation often ends up losing the original tone of the text.
Quality translation work requires extensive research, subjective interpretation of the text, cultural adaptations, and sometimes slightly more drastic changes. But it’s worth it, because it makes for a far more enjoyable experience.
Of course, there are cases where the debate is open and there is no objectively correct answer.
One such example is Japanese honorifics, which provide a lot of context on the relationships between people. Japanese honorifics are far more granular than what western languages generally offer. There, it is far more common for people of equal social standing, age, etc. to be on a last name basis and use -san, whereas in western cultures you would use first names, even when using the V form in languages with T–V distinction.
Some translators choose to drop honorifics entirely, because they are not translatable and require cultural knowledge to understand; but in my opinion they usually fail to complete the cultural adaptation by e.g. changing close friends to a first name basis, which makes little sense to me. Sometimes, the evolution of how people call each other is an important part of the story, and in ongoing serialized works, you just can’t predict this, so the safer choice is just to leave the honorifics as they are.
Any cultural adaptation requires careful consideration, but some are far more difficult to pull off than others. In my opinion, there is a fair balance between aggressively adapting everything, and leaving some of the original cultural context untouched for the reader or audience to absorb. Honorifics fit in the latter category: they are not difficult to learn and understand—there aren’t many of them, and their use is widely documented—and they provide too much information to be safely removed.
Sometimes, there are subtle details that can neither be translated nor preserved using only words. Using another example from Japanese, there are many different ways to say I. Some are markedly masculine and informal (ore), masculine but sometimes effeminate (boku), neutral and/or formal (watashi, washi), or feminine (atashi). Writing takes advantage of the nuance and sometimes has characters switch between those variations for specific effect.
Using those pronouns untranslated would be pretty clearly unacceptable. But instead of relying on translation notes in the margin or footnotes—which should only be a last resort—or worse, flow-breaking parentheses, consider using typography or vocal cues to your advantage here. There’s a million ways to do this, be creative.
Going back to “word for word” translation, it seems that style—which I consider lazy and cowardly—is prominent enough that some people are so used to it as to claim they “prefer translation to localization”. In my opinion, this is an ignorant stance that shows a complete misunderstanding of the translation process.
Part of this, however, originates from old anime dubs which didn’t actually bother translating the original script, but instead would dub a completely different story over the same images. Those hack jobs would go so far in their “cultural adaptation” that they would routinely show a complete mismatch between the script and the visuals, such as the infamous scene where Brock calls onigiri rice balls “jelly-filled donuts”.
For a very good exploration of this, and other aspects of anime translation, I would recommend this excellent video: Translation, Localization, Censorship, and You.
Finally, official vs fan-made translations.
For anime, the advent of simulcast has effectively eliminated low-quality fansubs. There isn’t really an incentive for fans to produce lazy translation jobs, because this is no longer the difference between a show only being available in Japanese or not.
Instead, the remaining fansub groups focus on doing a better job than the official subs. They put a lot more effort into research and script editing. They do karaoke and fancy typesetting. That’s because they are not limited by deadlines or budget, and it is a labor of love. There would be no reason for them to do this otherwise.
For manga, the sheer volume and far more limited budget of the industry means that only the most popular titles get licensed for an official translation, so fan work remains supreme. The quality is still hit-or-miss. Some groups do a stellar job, some people think OCR and machine translation is all they need, and in some of the worst cases, their English is piss-poor.
I don’t want to discourage anyone from trying, but a barely-edited machine-translated script is not “better than nothing”. Often enough, some people like to start drama about “sniping” when another group tries to do a better job, or—even worse—they encourage the low-effort translators who produce faster releases, and in turn discourage the quality-focused groups from continuing work on a series.
I don’t speak Japanese, but I think I would still be able to produce a quality scanlation using machine translation as a base, because I know that’s only a tool, one step of a process and not the end result. After all, a translator’s job is almost less important than an editor’s. I think it would be quite fun.
Anyway, please don’t be that idiot.
I think I’m out of things to say, and—as usual—I’m not good at writing conclusions.
Thanks for coming to my TED Talk.
XP-Pen sent me their new Artist Pro 16 (Gen2): they were curious if I could get it to work under Gnu/Linux without their proprietary driver.
Result: after a long evening of trial and error, tweaking Digimend/X11/xsetwacom, almost everything works now. 🎉
Of course, the deal is that if it works, I have to document it like I did with the Artist 24 Pro. So expect it soon™.
Btw, this one will stay on my desk: I had a serious crush on it while testing: super low latency & low parallax. 😍
Can we please have a setting in the administrator panel to set the character limit for a Mastodon instance?
The current lack of such a setting does nothing apart from create a two-tier fediverse where folks who have the time to maintain their own Mastodon fork and folks who use other ActivityPub servers can write as much as they like whereas everyone else is limited to 500 characters. We’re not kids; make the option accessible and let instance owners decide for themselves.