Translating GZDoom's text content. Read if you want to help

[Kostov]

Cool!
On the downside: A few more characters to draw - for nearly all fonts...

I'll do my best to take care of it. I at least owe you that much, not having helped you with the other stuff as much as I used to.

[Graf Zahl]

Well, it's just 5 characters more, but it's all ones that require some actual drawing. How come the Serbians have so many strange special characters? The ones for the other languages are nearly all just regular ones with accents.

[Xeotroid]

This table shows how much of a clusterfuck it is to try and get all of the different Cyrillic letters from all of the languages together. I guess in the past there was more Greek and Turkish influence on the language in southern Slavic countries, making their scripts evolve differently from those farther north? Hell, those five sounds in Serbian are also present in Macedonian, but two of them are just common Cyrillic letters with accents while Serbian went and did its own thing.

[Graf Zahl]

So what about Latin scripts, then?

The nordic languages have their Ø, while other countries use Ö for the same sound - Hungarians go one step further to add an Ő for a long version of the same sound.
Icelanders have their Þ and Ð, Germans have the ß, Chech uses the Ů - Esperanto invented its very own accented letters that aren't used anywhere else, and so on and so on.
The most hilarious thing is that in order to appease the Romanians, S and T with comma below had to be made separate characters from S and T with cedilla, although it's such a subtle thing that most people would never get the difference to begin with.
And that doesn't even consider the common accents and diacritical marks to be combined with base letters.

I find it truly baffling how inefficient some languages are for computing, and yet they remain as they are (well, good luck trying to teach most programs to handle Arabic or other languages where the style of the characters depends on context.) Those 5 Serbian letters are harmless by comparison, but it's a bit much for a single language.

Somehow the advent of computers has even made it HARDER to adjust a language to new situations, instead they try to preserve everything meticulously - 100 years ago people were more flexible in this regard.

[Xeotroid]

Somehow the advent of computers has even made it HARDER to adjust a language to new situations, instead they try to preserve everything meticulously - 100 years ago people were more flexible in this regard.

As far as I know, that's exactly how Þ (th) in English died. Typesetting machines imported from Germany didn't have the glyph, so the closest letter to it was used, forming the good ol' "ye olde". They don't look similar to me more than p does, but things might have been different in the Fraktur typeface or whatever was used. And then at some point they switched over to "th" completely.

However, that's still just a matter of a letter or two. I'm not sure what sort of reform you would need to make to completely change a language's or a culture's writing system to make it more computer-friendly – not just using a different font or merging a few symbols together, but also changing the order it's read in or the basic rules of how it works. A few hundred years ago with way lower literacy levels and people being quite disconnected from other parts of their country, this would have been sort of possible, albeit taking a long time, but nowadays, it would cause a mess, not just dislike.

[Graf Zahl]

Unfortunately you are correct. But with some languages the main problem is that even the printed form of the language took over all the weirdnesses from handwritten style, and those languages are the hardest to handle. Well, the end result will be that those languages may see poor support in software in general - I'd say most software can't deal with right-to-left well, for example. Or that stuff just looks butt-ugly (like Korean in GZDoom... )

[Xeotroid]

Yeah, I have no idea how Korean, Chinese and Japanese are supposed to work with low resolution fonts. Maybe Korean and Japanese kanas, sure, since there's a sane number of characters in these, but how the hell would Kanji get done? There's thousands of these and even a 32x32 sprite resolution would theoretically only allow for 1024 unique glyphs. I guess this thing speaks for itself.

[Graf Zahl]

There's thousands of these and even a 32x32 sprite resolution would theoretically only allow for 1024 unique glyphs.

Huh? You have 1024 PIXELS in such a glyph, resulting in 2^1024 = 1.797693134862315907729305190789e+308 possible characters with such a matrix.

Regarding input, AFAIK Chinese IMEs depend on Pinyin, but then I ultimately have to wonder why they don't switch to that fully - in order to work they'll need to be able to understand it anyway... But I guess in the end it comes down to the simple fact that culture is important, even when impractical.

On the other hand the Mongolians managed to replace their ancient script with Cyrillic during the 20th century and the Vietnamese with an extended Latin script, so such a thing is not impossible.

[Gez]

Hangul could work much more easily if you had a variant script in which the letters stand alone (like in most alphabets and abjads) instead of being combined into syllabic blocks. To take Wikipedia's example: ㄲㅜㄹ ㅂㅓㄹinstead of 꿀벌. Even if you included all the obsolete letters (the large majority of which are actually double letters, like ligature or how w was originally vv) you still get below 160 symbols to draw. That's okay.

However, with the block system, even with just the 24 letters from modern South Korean, you have thousands upon thousands of combinations.

[Xeotroid]

Oh, I'm dumb. Yeah.

Mongolian's switch to Cyrillic was forced for political reasons, if I recall correctly. However, there was a successful attempt to adopt Latin alphabet in the first half of 20th century, but that got reverted and replaced with Cyrillic. Looking up Vietnamese, the Latin alphabet was forced by the French colonisation, and it stuck. Makes me wonder what other places could have had attempts for this sort of change, like India or Thailand, both with their own unique scripts.

Edit: Speaking of Mongolian, there might also be the issue of "we need a change" vs. "if it works, it works", since Mongolian had way too many different alphabets over the years, so the population's stance on something more standardised may had been more favourable.

[Graf Zahl]

I think in both cases it came down to a highly increased rate of literacy after the script change as the reason for why it stuck. All those people who suddenly had access to literature would be very unwilling to give it up again, just for tradition's sake.

Such things are mostly decided by practicality reasons in the end.

[Kinsie]

Yeah, I have no idea how Korean, Chinese and Japanese are supposed to work with low resolution fonts. Maybe Korean and Japanese kanas, sure, since there's a sane number of characters in these, but how the hell would Kanji get done? There's thousands of these and even a 32x32 sprite resolution would theoretically only allow for 1024 unique glyphs. I guess this thing speaks for itself.

Kanji was largely avoided by Japanese game developers in the days before higher resolutions, for both legibility and memory reasons - it was a lot easier to fit 46 kana characters into your limited tile memory than 2136 kanji characters! (It was even easier to fit just 26 English characters, which is why many 80s Japanese games use poorly-written English instead of their native tongue).

[Captain J]

Done and done! i've just translated all the texts of Hacx! Tho not sure about the STSTR_MUS one... You may delete it clean if this one causes errors.

[Graf Zahl]

Why should it cause errors? My remark was about the content - all its meaning is lost if translated. But I cannot judge it for a language with an entirely different script.

[Kostov]

Funny thing about Serbian: many people online just write it with the Latin script and replace letters like č, š, ž with c, s, and z. I'm sure people would resort to that for unofficial localizations if some program doesn't support any extra characters, but it's a good thing we have Unicode.