Please add the Chinese language to the page of the website

Recently, I have time to slowly translate the words in the web page, please show the Chinese language, so that I can modify the content in time.
Also please modify Taiwan in the area. Please do not list Taiwan as a single country.

1 Like

Hi! We generally make translations available on the beta server ( once they have a certain percentage of content translated (I forget exactly what percentage that was, @yvanzo might remember). So if you translate for a while, we’ll make that available and after that it should be visible on the site to make sure it looks weird.

We’re not going to change Taiwan though, sorry. We’re fairly committed to democracy here :slight_smile:


Then there is nothing to be done. I’ll stop translating. In addition, Taiwan’s designation as a country is unacceptable in China and most other countries. So, please understand. Because this is a principle that involves the bottom line. Because of this problem, there is no way to open up to China. Because I believe that even if I translate all the words into simplified Chinese, this service will be abandoned due to national problems.

And on the question of democracy, I don’t think your ideas are democracy, because they don’t create a legally recognized state.

On the contrary, what you’re doing looks like a provocation.

Back to the question of countries, at least Taiwan is not yet an independent and recognized category of countries.

Thank you for helping with the translation to Chinese with simplified characters. I tried to add it to the test instance as a first step, but am still getting issues, probably related to the language code zh and zh_Hans. I will have to investigate it further.

Beyond political considerations, Taiwan is relevant as a release country anyway.


Does the United Nations recognize it?

No more. If the country’s problems can’t be solved, this can’t be opened.

In the same way, I can also say that what you are doing is not democratic at all, because you are classifying a region that is not officially recognized as a country.

I started working on the technical issue you reported and writing my above answer a long time before you started posting your very debatable replies. You seem to perfectly know that the status of Taiwan is a diplomatically tough topic, please check Wikipedia article for more information if it is available from your place; Teaser, it starts with “Taiwan, officially the Republic of China (ROC), is a country in East Asia.

As I mentioned above, it is mainly listed as a country because there are albums specifically released there. Many other area entities are also set as countries in MusicBrainz despite the fact that they are not recognized by the U.N. either.


Yes, I am well aware that this is a tricky political topic.
I would love to localize the site more, but the series of problems caused by this issue is not what I want to see.
But again, if you take the word “democracy,” what is democracy? It’s understandable what the people want.
So with 1.4 billion people on the mainland and 23.3 million in Taiwan, what do you call democracy?
Looking at this beyond political considerations, isn’t it at all undemocratic to classify an unrecognized region as a country?
Of course, I can also suggest that you erase the category of the Taiwan region, leaving only the Taiwan region, neither subordinate nor independent.

As for what you said about other entities listing Taiwan as a country. Because I look more at the i18n project is basically mature sites. So there is no such problem. This is the first one I’ve come across.

In addition, the wiki itself is available for anyone to edit. You can use it as a reference, but not as a source of authority.

If “democracy” is to be said, it is more appropriate to divide all similarly disputed and non-sovereign regions into a unified regional classification, which is neither affiliated to anyone nor an international with its own sovereignty. This is how “democracy” should be done.
Because they are neither incorporated into a sovereign state nor themselves a sovereign state, they can only be called “a certain region”.

Not sure if


As for what you said about other entities listing Taiwan as a country.

This is not what I wrote, I wrote that many other area entities are also set as countries in MusicBrainz despite the fact that they are not recognized by the U.N. either. Here are two examples:

In addition, the wiki itself is available for anyone to edit. You can use it as a reference, but not as a source of authority.

MusicBrainz itself is also a collaborative encyclopedia and is not a source of authority either, so it is perfectly relevant to refer to Wikipedia here.


I’d like to know if there’s a way to do language fallback, so that regional variations don’t need to be wholly translated, just as en-GB in Picard doesn’t contain every translated string from en-US where they can be identical.

Chinese (zh) is a macrolanguage. Wikipedia implements six regional variants (zh-CN, zh-TW, zh-HK, zh-MO, zh-SG, zh-MY) with a series of fallbacks:

zh-Hans > zh-CN > zh-SG > zh-MY
zh-Hant > zh-TW > zh-HK > zh-MO

On the one hand I think it’s an overkill to support eight scripts, but on the other hand, there are regional differences to some words.

I should add that the current standard seems to promote cmn (Mandarin) rather than zh (the macrolanguage of Chinese) to represent standard written Chinese. So maybe we should use that instead (e.g. cmn-CN, cmn-TW, etc.)

Edit: It will also be much easier to handle translations if zh-Hans and zh-Hant can reuse the same strings but have them go through an automated translator. With some exceptions, one simplified Chinese character can be translated/mapped to several traditional Chinese characters but not vice versa. So, zh-Hant is the more specific of the two can its strings can perhaps be translated automatically to zh-Hans through some library like OpenCC to save time and effort.


There is unfortunately no such mechanism right now. Projects that use gettext can have a bit this behavior, because it will e.g. fall back for a missing string in fr-CA to the string found in fr. But AFAIK this does not fully work for the Chinese variants (it probably would fall back to “zh”, but this does not make sense standalone).

And honestly the approach of only partially translating the language files like it is done in the English variants for Picard does not work well. It somewhat works for English as English is also the source language, so e.g. in the en-GB translation you just need to do the differences to the source. But I gave up trying this for e.g. the Swiss variations for the German translations. The differences in spelling are minor, but it is hard to find the strings that need changing and see which (new) strings need attention. It would be easier to one time translate all by copying the main language and then go over it to fix the differences.

What would help would be explicit support by tools like Weblate. There are some interesting ideas at Support fallback to the "generic" language variant · Issue #3412 · WeblateOrg/weblate · GitHub . So there is some hope we’ll get this one day.


Thanks @outsidecontext. What I fear is that if you have so many variants, the strings that can/should be the same become “de-synced” when some translator decides to improve one of them and doesn’t know about the other. It looks like there’s nothing much we can do about this.


As someone who has worked professionally in website/app localization, I think the de facto standard is to have two variants: zh-CN (simplified Chinese, Mainland) and zh-TW (traditional Chinese, Taiwan). If the content is originally Chinese, it generally comes from a Chinese company and they’ll just convert it with an automatic converter (often just Excel, actually), but still use -TW for it. If you’re a huge company, like Google, you may hire translators to work on regional Chinese variants, but that’s very rare. I’ve never seen cmn tag used, except to tag spoken Mandarin (e.g., a film may have a Mandarin (cmn) and a Cantonese track). And like the webpage you linked to says, “Using cmn or cmn-CN may cause serious compatibility problems if the software or users expect a tag such as zh.” I personally think it’s a mistake to call written standard Chinese Mandarin (though it’s of course much more influenced by Mandarin than any other topolect), but that’s debatable and this isn’t a linguistics forum.

I see the translations list has the -Hant and -Hans tags, this should probably be changed to -TW and -CN. Firefox uses these tags for the request-language header, presumably other browsers use the same ones (CN TW), if MB used -Hant and -Hans instead the browser may pick up the wrong variant.