I have a question about the count of hosted words in localization with Japanese as the original language.
I know that hieroglyphs are counted as one word per character.
No matter what word is set for the translation language, the usage shown in the URL below is converted based on the number of hieroglyphic characters.
On the other hand, if you enter hieroglyphs in the target language, the count of hosted words is based on English words only, not hieroglyphs entered in the target language.
In other words, is it possible to use words with low word counts as source words in crowdin in order to reduce the number of hosted words?
I am wondering if I should use English as the source language because the amount of Japanese words in the source language would be too large.
Answering directly, hosted words means number of words x languages, so if you have 100 words in source files, and 3 target languages, number of hosted words would be = 300.
Consider it from different point of view. Like, what would be you workflow, or so. As far as I know, language pairs that contains English are more popular (and cheaper) within freelance translators than Japanese. Plus more Machine engine supports English as a source.
On the other hand, if you already have source in Japanese, you can definitely consider it’s as a primary language.
I guess this is the question no one can provide undoubted answer, both ways have pros and cons.
To answer your question directly, the number of hosted words is the number of words x the number of languages, so if you have 100 words in the source file and 3 target languages, the number of hosted words = 300.
Yes, that is correct.
However, I was a little concerned about this word count because it depends on the language.
Because the word count is the number of words x the number of languages, the hosted word count should not exceed 200, even if the content is as follows
English (original) 100 words
Japanese 300 words
French 120 words
On the other hand, if the original text is in Japanese, the number of words is 600.
Since even a single word makes a big difference, the longer the sentence, the bigger the difference seems to be.
For example, apple is one word in English, but three words in Japanese.
My project is in Japanese.
In this case, to save on word count, it seems cost-effective to convert it to English once and then translate it into each language using English as the source text
Hello there, seems logical. Never though about it, but seems logical. I guess by default it’s better to use English as source, because of Crowdin Machine Translation Engine - it’s totally free of charge, but it works with English as source only (maybe with some languages, but I don’t think there’s Japanese in that list since it’s quite complicated), so on the long term distance, you’ll save time and efforts using Crowdin’s Machine Engine.
@Tatsuhito_Usui it very depends on style of writing. I know that for example, both Crowdin and MS Word would recognize different script styles (like Katakana and Hiragana) diffently. I’ve heard that one of those styles is recognized pretty similar to English, but I can’t remember which one.
Anyone, giving +1 to English sources because of Translation Engine
I’ve jumped today here in chat and decided to join this thread as well
Maybe your domain of work or project can be considered as open source, or particular charity related, or educational or so? I know Crowdin offers nice discounts for all who support other people around the world:
Scroll into the end of page, there would be open source + educational notice :
