Conversation
Per-player language support for quest markdown files is not yet implemented.
|
I have this as a draft right now because I haven't actually added any translations, and because I haven't done anything yet about the questbook markdown content parsing. |
TheEt1234
left a comment
TheEt1234
left a comment
There was a problem hiding this comment.
are you sure that you like doing these kinds of changes? this is a lot of repetitive work for not much in return imo
translations may make it harder to change sbz, and i don't know if it would be possible to verify if they are accurate/good
i think a bad translation < english, maybe i am biased
| @@ -0,0 +1,413 @@ | |||
| # SOME DESCRIPTIVE TITLE. | |||
| # Copyright (C) YEAR THE PACKAGE'S COPYRIGHT HOLDER | |||
There was a problem hiding this comment.
Yeah, that's literally just what gets generated by the gettext. I just had to get some sleep, so I left it that way for now.
| if #new_desc > 1 then | ||
| for i = 2, #new_desc do | ||
| new_desc[i] = core.colorize("#777", new_desc[i]) | ||
| if not new_desc[i]:find('\027') then |
There was a problem hiding this comment.
Why? and does this work properly
There was a problem hiding this comment.
Yeah, there were errors popping up because this was trying to colorize translated text that was already colorized. I sort of naively just made a condition to avoid doing that. This is an instance where I think you're right in that it's going to be somewhat more complicated to have translation support in the mix.
There was a problem hiding this comment.
Yeah, there were errors popping up because this was trying to colorize translated text that was already colorized. I sort of naively just made a condition to avoid doing that.
oh, ok (i thought this was checking for checking if there are any translations and not coloring if there are)
There was a problem hiding this comment.
Nah, you can still color translated text. You kind of want to get as close as possible to the pure strings when you wrap with S() and PS() though, I believe. You can have translations with format specifiers embedded in them, but I imagine it can start to get weird and unnecessary at some point.
mods/sbz_logic/locale/sbz_logic.pot
Outdated
| # SOME DESCRIPTIVE TITLE. | ||
| # Copyright (C) YEAR THE PACKAGE'S COPYRIGHT HOLDER | ||
| # This file is distributed under the same license as the PACKAGE package. | ||
| # FIRST AUTHOR <EMAIL@ADDRESS>, YEAR. |
There was a problem hiding this comment.
oh i see its for all the files, you should probably change these comments idk translation stuff
There was a problem hiding this comment.
Yep! I'll need to be a little careful though and reference the copyright mds.
mods/sbz_logic/knowledge.lua
Outdated
| @@ -1,3 +1,5 @@ | |||
| local S = core.get_translator(core.get_current_modname()) | |||
There was a problem hiding this comment.
I would prefer if you didn't touch sbz_logic or sbz_logic_devices as i would get lots of merge conflicts from this (im completely rewriting it)
but its fine, you can do it
There was a problem hiding this comment.
Sure thing. That makes sense.
Yeah, I'm sure. I imagine it's like those people who make ships in a bottle feel. It's so tedious and precise, and there's so little possible fanfare at the end of it that it seems like you'd have to be not right in the head to do it. But I like it. I like knowing that I've contributed in a way that might not otherwise ever happen (or at least not to the degree I want to push toward). When I believe in something like this, I want other people to see it in its best light. Knowing that I've done everything I can is enough for me to be able to walk away and feel proud. Doesn't matter if other people see it. I see it. That being said, there's still burn-out to contend with. If this were an actual job where I got paid and was expected to push through, I would have to. Like with the sound design that I'm taking a break from currently. But I have the luxury of largely choosing whatever I want to work on at any given time. So, I'm okay. |
|
Regarding translation support being harder, I think it's just bitter medicine (like forcing Zander to learn git 😜). It's going to widen the player base though and make some of the existing players more comfortable. This is largely just me virtue signaling at the moment, but I really do think localization and accessibility should be closer to the top in importance on all projects. I'd like to meet the challenge of it. Verifying good translation will be difficult, but there are already methods of doing a first pass on that, like having anyone familiar with the language give it a once-over or translating it back to English with a translation engine and looking for stuff that sounds wacky. I think as long as we're not intentionally or unintentionally making these translations as bad as they can possibly be though, we don't have a very high bar to cross. imo, it will be better to have something as opposed to nothing. If we don't even start, we aren't going to get people chiming in about fixes. And so long as we get the process done thoroughly and thoughtfully enough, the fix shouldn't have to be "scrap the whole thing and start over". But if it is, that's something we'll learn from. And I would want to. Because I think it's important. |
|
Maybe having a Weblate would be good as well? I've seen it on other projects. |
Like something to automatically create the translations as a continuous integration type of thing? I don't want to do anything too hands off, but I'm not opposed to using generated translations as very-first step to touch up on and quality check thoroughly after. |
|
idea: we could use weblate and codeberg https://translate.codeberg.org/ |
|
i like how the aes luanti server is listed first in there lol |
|
Oh, so weblate is like an interface for contributing? |
|
I think so |
|
Super hot take, but I would prefer not leaving major languages mostly or entirely untranslated in hopes that some day someone comes along wanting to contribute anything. We have technology to speed things along to a point at which it can be of a serviceable quality, even if things here and there aren't 100% good or 100% accurate to what was intended in the source language. I would absolutely love to have full translations by human beings with good firsthand knowledge of the to and from languages; I just don't know that we can expect that in every case. And I think it disenfranchises people worse to not have anything or just scraps here and there for months to years. The weblate interface looks great, but it sort of takes a stand about maintenance and quality assurance that we as developers on this project I don't think have all signed up for. We seem to like things pretty good. Correct me if I'm wrong here. Because if we have the manpower and willingness to back up a serious attempt at QA on translations, I would be extatic. Otherwise, I would recommend we settle for as good as we can do with what people and resources are available to us (until more comes along later, hopefully). |
Doing this because lots of stuff is moving and shifting over on frog's fork.
Doing this because lots of stuff is moving and shifting over on frog's fork.
Extracts parse_md_file as a private helper that returns quest data rather than registering it directly, making both easier to extend later. It'll just fall back to English on missing file instead of crashing.
I think we need to do this so that it's clear to the translators that this is not to be changed, and they can still have the localized titles as markdown titles. I'm not sure if the localization files need to worry about requirements though, since that should probably be a single source of truth in the en version.
3 participants
I've gone through the game looking for player-facing strings (item names/descriptions and chat messages mostly), and I've wrapped those in S() and PS() where applicable. I've also added locale directories and a script to create translation templates from all the occurrences of those wrappers.