Template talk:Nihongo/Archive 4

Archive 1

Archive 2

Archive 3

Archive 4

Archive 5

Using Nihongo template and zh template together

On the page Tungchow Mutiny the Nihongo and Zh templates are used together. This creates inconsistent punctuation. The items on the zh tempate are separated by colons, the nihongo template has no separator, but the junction between the two templates is marked with a comma. For example:

• The {{nihongo|'''Tungchow Mutiny'''|通州事件|Tsushu jiken|{{zh|c={{linktext|通|州|事|件}}|p=Tōngzhōu Shìjiàn}}|lead=yes}}, sometimes...
  

The Tungchow Mutiny (Japanese: 通州事件, Hepburn: Tsushu jiken, Chinese: 通州事件; pinyin: Tōngzhōu Shìjiàn), sometimes...

Can we have consistency between the two templates so that they can be used together neatly. Rincewind42 (talk) 06:42, 1 January 2014 (UTC)

That's not an appropriate way to use this template. The better solution is to use Template:CJKV. I'll go ahead and edit Tungchow Mutiny to do so, since no one else has done so in the intervening 4 years. --Bigpeteb (talk) 16:57, 16 March 2018 (UTC)

New name for "Help:Installing Japanese character sets"

Recently I did some work on the page Help:Installing Japanese character sets, to make it clear that the page is about installing fonts. Users don't install character sets (like Unicode, an international standard), they install fonts and/or make various other settings so that they can display and enter characters in the character sets. So now I would like to move the page to a new name. Two questions:

1. Would it be possible to update the nihongo code so that it goes to the new page when a user clicks the question mark? I don't want to rename the page if it means that everyone is going to arrive via a redirect.
2. Does anyone have any good ideas for a new name?

Some ideas that occur to me are

• Help:Installing Japanese language fonts
• Help:Displaying Japanese characters
• Help:Japanese language support

Does anyone have better ideas for a new name? There are other pages about this on Wikipedia, so maybe we should try to choose a consistent name. Thanks. --Margin1522 (talk) 06:07, 1 September 2014 (UTC)

Looking though at the equivalent for Chinese and it covers Japanese too: Help:Multilingual support (East Asian). It goes into more detail as to what to expect then the detail for different OSes is largely the same. It probably makes sense to merge the two pages then link to the combined page. This is consistent with other pages at Category:Wikipedia multilingual support which includes similar pages for e.g. Indic languages, rather than pages for individual languages.--JohnBlackburne^words_deeds 17:57, 1 September 2014 (UTC)

That would solve the renaming problem. I would be open to that. I like the support test with images. One problem is what to do with the explanation of the nihongo template. Ideally we would want corresponding sections for Chinese, Korean, and Vietnamese, but I don't know enough about them to write it. If someone could do that, I think it would be useful to have simple explanations of all 4 together on 1 page, with links to the detailed pages. Although that should probably be a separate article.

Another issue is the imbalance in hits. Currently Help:Installing Japanese character sets gets well over 1000 hits per day, while Help:Multilingual support (East Asian) gets about 200. I suspect that a lot of those 1000 hits are from people who can see the characters fine but click the question mark to see what that's about. Since they came, we want something to show them, which is the rationale for the template explanation.

A final issue is that neither article has anything about mobile. My impression is that right now mobile phone users are the ones experiencing the most problems with Japanese. E.g. see this [reddit thead]. I think we should recruit someone who could do that. I can't because my only experience is with Japanese operating systems, and I've never had any problems. --Margin1522 (talk) 19:22, 2 September 2014 (UTC)

ja-Latn-alalc97

Hello,

Should we add {{lang|ja-Latn-alalc97|{{{3}}}}} to the template? Last versions of web browsers don't change the font if Latn subtag is used, except Firefox, it will be fixed on version 33 out October 14.

alalc97 = Modified Hepburn method [1]

Sorry about my English Thibaut120094 (talk) 14:29, 9 October 2014 (UTC)

Good to see this at last being fixed in Firefox. The Latn tag is correct and strongly recommended for [Latin] Romanisations, and is supported by [all?] other browsers. The subtag alalc97 is also correct though less important; no browsers that I know support these but users can make use of them using custom CSS: [2].--JohnBlackburne^words_deeds 15:01, 9 October 2014 (UTC)

Well, since they are different Romanization methods (including Traditional Hepburn), I think it's important to add the subtag for the modified Hepburn method, so people who use speech synthesizers can correctly hear the transliteration with the correct method and pronunciation. Chrome/IE on Windows 8.1 and Safari on iOS don't change the font, I don't know for others browsers / platforms though. Thibaut120094 (talk) 15:12, 9 October 2014 (UTC)

Yes, such subtags can also be used by speech synthesisers. Any user can enable styling based on language using CSS; I do so here: User:JohnBlackburne/common.css. This was when working on Template:Zh and its associated module and making sure all the support in that worked. That template now has fairly complete support for tags, script tags and variant subtags and apart from a few users experiencing problems with Firefox (see Module talk:Zh#Font?) there have been no real issues.--JohnBlackburne^words_deeds 15:38, 9 October 2014 (UTC)

I've added it to the sandbox {{Nihongo/sandbox}}. Have a look at the testcases, Template:Nihongo/testcases, in particular the third column which uses the sandbox version. It's working as best I can see – due to my custom CSS the 'Nippon' (and Nihon) throughout is coloured indicating it's tagged with a language tag. That's with Safari on Mac OS.--JohnBlackburne^words_deeds 15:50, 9 October 2014 (UTC)

Edit request

Please update the main template from the sandbox with the change above. This is a straightforward and I think uncontroversial change, similar to changes made to Template:Zh recently, and correctly tags the Romanisation in a similar way. The only issue that arose with that was with the Firefox browser and as noted above that should now be fixed. I was hoping to get some feedback on it but this has been sitting here for over a month without any comment, and my recent post to WT:JA did not elicit a response. I can only assume anyone looking at it found it unproblematic. Probably the only way to find out if there are any issues is implement the change and look out for problems in articles and reports in the obvious places.--JohnBlackburne^words_deeds 01:53, 26 November 2014 (UTC)

Done Let me know if there are any issues with it. — Mr. Stradivarius ^{♪ talk ♪} 06:05, 26 November 2014 (UTC)

I have Waterfox 33.0.2 and the romaji field now looks strange.--十八 04:10, 27 November 2014 (UTC)

Looks a bit odd to me too; the font is different from that used for English in Waterfox (which I've not tried before) and Firefox.Template:Zh/testcases looks fine though. Lets try some tests with Chinese and two ways of doing Japanese romanisation. Look OK in Safari; the Chinese romanisation is OK in Firefox but the Japanese one's aren't. Odd.--JohnBlackburne^words_deeds 04:26, 27 November 2014 (UTC)

• {{lang|zh-Latn|Zhongguo}}: Zhongguo
• {{lang|ja-Latn|Nippon}}: Nippon
• {{lang|ja-Latn-alalc97|Nippon}}: [Nippon] Error: {{Lang}}: unrecognized variant: alalc97 for code-script pair: ja-latn (help)

Ack no no no no revert this immediately. The whole compliment of -Latn tags do not work. They encode the romanized text as if it was Japanese text. It looks god awful italicized. Mr. Stradivarius Change it back.—Ryūlóng (琉竜) 05:53, 27 November 2014 (UTC)

Already done by SiBr4. — Mr. Stradivarius ^{♪ talk ♪} 08:46, 27 November 2014 (UTC)

@Ryulong: By the way, your ping didn't work - I was notified about this because my edit to the template was reverted. Not sure why it didn't work, unless you copied and pasted part of your signature. — Mr. Stradivarius ^{♪ talk ♪} 08:50, 27 November 2014 (UTC)

I thought {{u}} worked.—Ryūlóng (琉竜) 08:51, 27 November 2014 (UTC)

Yeah, this doesn't work. As explained in the last comment of the FF bugreport: "Unduping: bug 756022: the fix for that was narrower in scope than this bug and didn't address the issue of script subtags in language tags in content." So they fixed this for some parts of Firefox, just not for random pieces of content with a script subtag in the lang attribute. —TheDJ (talk • contribs) 13:03, 27 November 2014 (UTC)

Ryulong and anyone else with problems, can you please be more specific about what you're seeing and with which browser(s) (and which versions). If it's as I suspect a Firefox font problem then it can be brought to their attention as still not fixed, contrary to the report at the top of this section. It should work, it's the proper way to mark languages, and as far as I know all other browsers handle this properly.

One point to make is it may depend not only on using Firefox but on the settings used, At least this was the case with {{zh}} where the only people with problems were those who'd changed the fonts for Chinese to other than the defaults. It should not have made a difference as for text marked with Latn it should use Latin e.g. English fonts not the fonts specified for Chinese. Looking at the settings here it may be that Japanese is different, i.e. worse, as the default fonts for Japanese are not 'Serif' and 'Sans-serif' but Japanese specific fonts, but this may vary depending on your own setup.--JohnBlackburne^words_deeds 13:21, 27 November 2014 (UTC)

Ah thanks TheDJ, So still not fixed. And a bug that was first reported over a decade ago. I would not hold out any hope that they'll get around to it any time soon.--JohnBlackburne^words_deeds 13:26, 27 November 2014 (UTC)

For information, it's now fixed on Firefox 36. Regards, Thibaut120094 (talk) 20:33, 3 March 2015 (UTC)

Revisit

It's been more than 3 years. It appears that browsers have improved and handle language script and variant tags much better now. Is it time to try adding this tag again? --Bigpeteb (talk) 17:38, 16 March 2018 (UTC)

Placement of "Help:Installing Japanese character sets"

I would like to know why the "Help:Installing Japanese character sets" link ("?") typically shows up after the Hepburn and not the Japanese (kanji/kana). It seems to me that it would make more sense putting it directly after the latter. – Maky ^« talk » 19:26, 22 November 2014 (UTC)

Strange error, not sure if it was discussed already

The current formatting of this template introduces a problem in cases where it is used to mark a term that is already written in romanized Japanese, doubly so when used for the name of a Japanese person written in western order.

The article on Akira Toriyama currently begins "Akira Toriyama (Japanese: 鳥山 明, Hepburn: Toriyama Akira)". The fact is that his name is in Japanese in all three instances, and both romanized spellings are Hepburn. What the text actually means here is "Akira Toriyama (kanji/kana: 鳥山 明, Japanese naming order: Toriyama Akira". As far as I can tell this problem extends to thousands (?) of other articles on Japanese people born after 1868, and a lesser version to all the other Japanese bio articles before 1868.

I can think of three possible solutions:

1. A new template is made for Japanese personal names and replaces this one in all those instances.
2. This template is reverted to its previous form without the explanation of what each bit means. This seems like the best solution to me right now, but I didn't read the discussion that led to the change.
3. New wording is made to accommodate all eventualities, such as "Akira Toriyama (Japanese script: 鳥山 明, direct transcription of Japanese script: Toriyama Akira)". This is ... obviously not desirable, but maybe someone else can imagine less awkward wording.

Am I missing something? Was this problem already given a technical fix I haven't seen, and the Toriyama article hasn't been fixed yet?

Hijiri 88 (聖やや) 14:01, 28 December 2014 (UTC)

The "Hepburn:" is there to give readers something to click on to find out which romanization system is being used to transcribe the kanji. It's a pronunciation guide. If you don't want the "Hepburn:", you can leave out the "lead=yes" parameter. About the "English" parameter, people are free to use any system they like for their own names. Especially if the name has long vowels, they are likely to use o, oh, ou, or oo instead of ō. So they may be different. About the order of the names, if you want you can use {{Japanese name}}, which gives

In this Japanese name, the surname is Toriyama.

– Margin1522 (talk) 14:49, 28 December 2014 (UTC)

If Lead=yes for the question mark

Should it be encased inside the if lead clause? I think just having it appear once is enough. DragonZero (Talk · Contribs) 05:15, 31 December 2014 (UTC)

Arguably, people should use {{nihongo4}} (which doesn't have the question mark) after the first use of {{nihongo}}. I've been trying to fix that on various pages I've come across as I see it.—Ryūlóng (琉竜) 23:53, 19 January 2015 (UTC)

Adjusting the current template would be easier and more accessible to new users, since lead will also only be used once. DragonZero (Talk · Contribs) 04:16, 20 January 2015 (UTC)

Lead isn't used that often as a parameter. It's easier to just go on pages, see if they use nihongo4 or not and impliment it with a find and replace text thing.—Ryūlóng (琉竜) 04:34, 20 January 2015 (UTC)

It should be. It's to increase accessibility to those unfamiliar with Japanese articles. I'd like to put this up to RFC for a straw vote if a consensus can't be reached. DragonZero (Talk · Contribs) 02:35, 31 January 2015 (UTC)

Related to this, I think, is the unanswered question I raised above about whether the "?" should follow the Hepburn or the Japanese (kanji/kana). It seems to me like it's backwards since the Hepburn will show up anyway. If people need Japanese character sets, they'll noticed with the kanji/kana. – Maky ^« talk » 03:24, 31 January 2015 (UTC)

Yeah, that was going to be my second suggestion after this. I tried playing with the code to move it but I've failed. DragonZero (Talk · Contribs) 04:50, 31 January 2015 (UTC)

Template-protected edit request on 6 February 2015

The following discussion is an archived record of a request for comment. Please do not modify it. No further edits should be made to this discussion.A summary of the conclusions reached follows.

Yes and no respectively, appears already to have been actioned. Guy (Help!) 10:56, 17 April 2015 (UTC)

---

• Move the Question mark to the Japanese characters. This can be skipped if too difficult.
• Put the Question mark inside a lead clause, so it would only show up with (|lead=yes). I have done this in the sandbox already.

It shouldn't be anything controversial but could be settled by a straw vote. If Template size is a worry, users can just opt to use Nihongo4 instead. DragonZero (Talk · Contribs) 06:38, 6 February 2015 (UTC)

I'll need to see consensus for these changes. Please reactivate when needed. — Martin (MSGJ · talk) 09:42, 6 February 2015 (UTC)

RFC February 9, 2015

Needs consensus on proposal 1 (Move question mark to Japanese characters) and proposal 2 (Put question mark inside a lead=yes clause). DragonZero (Talk · Contribs) 02:11, 10 February 2015 (UTC)

Proposal 1

Survey

• Yes I think it would make more sense having it hover over the potential squares instead of the romaji. DragonZero (Talk · Contribs) 02:11, 10 February 2015 (UTC)

• Yes This seems a no-brainer, but see below. Imaginatorium (talk) 06:30, 10 February 2015 (UTC)

• Yes As the person who proposed it. – Maky ^« talk » 07:05, 10 February 2015 (UTC)

• Definite yes. It absolutely belongs with the Japanese text, not the Romaji. VanIsaac_WS^cont 09:15, 12 February 2015 (UTC)

• Yes I've wondered why the question mark was where it was for a while now actually. —KirtZ^Message 01:05, 16 February 2015 (UTC)

• Yes, provisionally Although my preference is to get rid of the question mark altogether and use {{Contains Japanese text}} instead. That is easier for users to understand. (Changing my vote, after thinking about it. I for one find the question mark very annoying. So no, I can't support displaying this right next to every single kanji. It makes sense, but I will support it only if Proposal 2 passes, to display it only once.) – Margin1522 (talk) 22:04, 19 February 2015 (UTC)

Discussion

Proposal 2

Survey

• Yes Having it appear once is good enough. DragonZero (Talk · Contribs) 02:11, 10 February 2015 (UTC)
• Unsure Do we know how many people the question mark actually helps? Is it really needed at all? I only ever encounter "boxes" in very obscure languages (I typically know nothing about), though it may help that I use Linux. I am not convinced that restricting this to the lead is a good idea: if anything, if Japanese appears in the lead, it is likely that the article is about a Japanese topic, and thus likely that the user is accustomed to reading Japanese topics, and sorted this all out years ago. On the contrary, the place where Japanese characters are unexpected is where some term or other quite peripheral to the actual subject is glossed with its Japanese origin (which I think is an extremely good idea); this is likely to be not in the lead, and to be read by users who do not normally read Japanese. (I can't immediately think of a good example, sorry.) Imaginatorium (talk) 06:36, 10 February 2015 (UTC)
  • About how many people click it, Help:Installing Japanese character sets is linked to by {{Contains Japanese text}} and the question mark, and gets about 1180 hits/day. By way of comparison Help:Multilingual_support_(East_Asian) is linked to by {{ChineseText}} and gets about 240 hits/day. So it looks like at least 900 users click the question mark every day. – Margin1522 (talk) 21:30, 19 February 2015 (UTC)
    • That helps, but unfortunately we don't know how many people clicked the question mark because they needed help installing Japanese character sets, and how many clicked because they wondered what on earth it meant. Imaginatorium (talk)
• Unsure per Imaginatorium. One example I'm wondering about is found at Love#Japanese. – Maky ^« talk » 07:07, 10 February 2015 (UTC)

Good example: and it shows another problem. The 愛 appears with a question mark in the "Japanese" section of this page, but has already(!) appeared above in the Chinese section, with no question mark. This suggests to me that the whole thing needs rethinking. Imaginatorium (talk) 07:34, 10 February 2015 (UTC)

Because it has both Chinese and Japanese I've added a {{Contains special characters}} to the article. This sort of makes the '?' next to the Japanese a bit redundant, but it's not causing any problems. As a technical matter although the same character 愛 is not really the same word in Chinese and Japanese, no more than 'fort', say, is the same word in English and French.--JohnBlackburne^words_deeds 01:59, 16 February 2015 (UTC)

Well, "fort" comes from "fortress", which means "strong-hold", and French "fort" means "strong", so this not a good example. (I think you meant like "hurt=pain=bread".) But actually 愛 is the same character (which is what matters here), but also has more or less the same meaning, and even almost the same pronunciation. Imaginatorium (talk) 04:23, 16 February 2015 (UTC)

Actually what I think we need here is a new "Contains East Asian characters" template that would link to Help:Multilingual_support_(East_Asian). {{Contains special characters}} links to Help:Special characters, which doesn't say anything about Chinese or Japanese.– Margin1522 (talk) 22:12, 19 February 2015 (UTC)

• oppose This would cause wholesale change in the majority of uses of the template and so should not be done without a very good reason which I don't see here. If the intention is to add a toggle for a behaviour then it should be added but default to 'on', i.e. the current behaviour, with way added to turn it off. But as noted above this already exists in {{nihongo4}}, so nothing else needs to be done.--JohnBlackburne^words_deeds 11:07, 10 February 2015 (UTC)
• oppose. I don't think it can be justified to tie in the appearance of the question mark with the lead parameter. Not all articles will necessarily have the first instance of this template in the lead, and this info needs to be available in any article in which the template is used. VanIsaac_WS^cont 09:35, 12 February 2015 (UTC)

Discussion

The majority of articles that use Nihongo with all three fields filled should have lead=yes on its first instance, whether it's in the lead or not. Secondly, we're getting into the weird practice of using Nihongo once (Lead or not lead=yes), and nihongo 4 for the rest. About the love article, since it's just the romaji and character, it should probably use nihongo3. DragonZero (Talk · Contribs) 19:57, 10 February 2015 (UTC)

Maybe I'm just lazy, but can someone succinctly explain to me why we need multiple Nihongo templates? (I'm deliberately not reading the template documentation here to simulate the confusion a new editor would experience.) Why can't they be condensed into one general template with several parameters to meet everyone's needs? – Maky ^« talk » 10:06, 16 February 2015 (UTC)

• Nihongo2 is for the Japanese characters only, question mark off by default
• Nihongo3 is when the Romaji is placed infront for things without official translations
• Nihongo4 is no question mark
• Nihongo-S seems to be the same as Nihongo4.
• Nihongo Foot is for placing the Japanese characters and romaji in the footnotes. DragonZero (Talk · Contribs) 19:23, 16 February 2015 (UTC)

It seems to me like Nihongo, Nihongo2, Nihongo3, Nihongo4, and Nihongo-S should be merged into a more versatile Nihongo template and the others deprecated. Additional parameters could be added to Nihongo to improve functionality and allow the flexibility that you seem to want with Proposal 2. Am I missing something obvious and horribly wrong? – Maky ^« talk » 08:28, 18 February 2015 (UTC)

Nihongo is already the Swiss Army knife of language templates. We have these others because nobody can remember the order of the parameters, so it takes two or three tries to get it right. – Margin1522 (talk) 22:03, 18 February 2015 (UTC)

Then we just need to better document the template so anyone can figure it out. I agree with what others have said here: there are too many and they should be combined into one. ···日本穣^? · 投稿 · Talk to Nihonjoe · Join WP Japan! 17:17, 19 February 2015 (UTC)

Nihongo-S should NOT be merged. It is meant to be a slimmed down version of nihongo for large lists where the nihongo template breaks do to going over the post‐expand include size. Merging it defeats that purpose. —Farix (t | c) 18:58, 19 February 2015 (UTC)

This might be able to be alleviated by using Lua, though I don't know how to use it. I know it's been used to reduce the footprint and processing time of other templates. ···日本穣^? · 投稿 · Talk to Nihonjoe · Join WP Japan! 06:55, 20 February 2015 (UTC)

Wow, I just looked at it and with Lua, size won't be a problem. That'd made Nihongo-S pretty obsolete.. DragonZero (Talk · Contribs) 08:16, 20 February 2015 (UTC)

I re-implimented the similar {{zh}} in Lua and it had many benefits. It had faster, cleaner and easier to understand code, made it easy to fix various longstanding bugs and issues, and easy to add new features. It also made it possible to merge other templates into it, in particular the former {{zh-full}} (see its documentation), implemented as a separate template due to limitations of parser functions, could be merged with all its sub templates. The same would be possible here, and probably quite straightforward as it's simpler template.--JohnBlackburne^words_deeds 10:34, 20 February 2015 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. No further edits should be made to this discussion.

Combining the Nihongo templates

In response to Nihonjoe's opinion that we should combine the templates, I'd like to suggest that if we do that we consider a radical simplification, with the goals of reducing the number of templates, making the HTML more efficient, and (most importantly) making the basic Nihongo template less complex for the user. I did some counting, and here are some numbers. To avoid clogging up the survey, I'll start a new section. The numbers in parentheses are the HTML output.

• Nihongo: Tokyo (東京, tōkyō) (439 chars, 6 spans)
• Nihongo2: 東京 (73 chars, 2 spans)
• Nihongo3: tōkyō (東京, Tokyo) (368 chars, 4 spans)
• Nihongo4: Tokyo (東京, tōkyō) (184 chars, 3 spans)
• Nihongo-s: Tokyo (東京 tōkyō) (66 chars, 1 span)
• Nihongo foot: Tokyo (103 chars, 2 spans)

References

In terms of the HTML output, they all wrap the kanji in (lang="ja" xml:lang="ja") spans. That looks like the baseline essential function. The largest are Nihongo and Nihongo 3, because of the code for the question mark. Some are larger because of <span class="t_nihongo_kanji">. I'm not sure what that does. Nihongo-s doesn't have it, and seems to display fine for me.

In terms of complexity for the user, a lot of the complexity comes from having the English as the first parameter, even though it is displayed outside the parentheses in every template except Nihongo3. IMO, if the English is going to be outside, the user can just write the English first. There is no need for this to be a parameter.

• Nihongo3 could easily be written like this: Tōkyō ({{Nihongo2|東京}} Tokyo). So I see no special need for Nihongo3.
• Nihongo foot could easily be written like this: <ref>{{Nihongo2|東京}} Tōkyō</ref> So I see no special need for Nihongo foot either. The only thing it does is save a few keystrokes.

So it looks to me like there is an opportunity to make this much simpler and easier for the user, plus more efficient HTML. Get rid of the English parameter. Base the code on Nihongo-s (with class="t_nihongo_kanji" if that's doing something essential). For Nihongo have the kanji be the first parameter, romanization as the optional second parameter, and all of the other options as named parameters that don't depend on the order. Then also keep Nihongo2 for the times that all you want is the kanji. Also perhaps keep Nihongo-s for the times when code size is important and options are not. – Margin1522 (talk) 21:50, 19 February 2015 (UTC)

I agree that it would be a good idea to simplify and reorganise. But I think the whole thing needs extensive thought: there are a number of quite different ways in which this is used, and it would be better to standardise the formats for these different ways of use. In particular, the "English" parameter is extremely ill-defined: does it mean "The English word for which this is a translation", "The (direct) English translation of a Japanese term being explained", "The way the name is written in materials at least somewhat in English", etc etc. These should be distinguished, and as far as possible presented in a standard way. Imaginatorium (talk) 04:29, 20 February 2015 (UTC)

If the English is not used in the template, then what's the point of it? Why not just use {{lang|ja}} (and {{transl|ja}} for the rōmaji)? Curly Turkey ¡gobble! 05:16, 20 February 2015 (UTC)

I actually happen to like having the English Kanji Romaji encasement, so I can not agree with this proposal. DragonZero (Talk · Contribs) 06:16, 20 February 2015 (UTC)

The English is used a couple different ways: direct translation and official translation. Both are valid uses as the field is meant to house English. Romaji is not usually the same as English, though it can be in rare cases (and in those cases, it should be left blank as it would be redundant). I agree that there really is no reason to have Nihongo foot. All the others were created by people who wanted to have the template display things in different orders. This might be able to be accomplished using switches in the template, or perhaps someone familiar with Lua (I think that's what it's called) can whip up something which could make things smaller in the template itself. This definitely should be rushed into, especially given how much some of these templates are used. ···日本穣^? · 投稿 · Talk to Nihonjoe · Join WP Japan! 06:33, 20 February 2015 (UTC)

Curley Turkey: If the English is not used in the template, then what's the point of it? Unless I misunderstand something here, from the point of view of a reader, only Nihongo3 uses [what I'll call for simplicity's sake] English. The article Tsuguhara Foujita starts {{nihongo|'''Léonard Tsugouharu Foujita'''|藤田 嗣治|Fujita Tsuguharu|November 27, 1886 – January 29, 1968}}, but I don't think that the the result implies anything about the status of the string "Léonard Tsugouharu Foujita" that would not be inferred if a template weren't used. ¶ As for [what is now] Nihongo3, I agree with Imaginatorium: a template that does appear to present "the English" for ("the Western version" of?) a Japanese term should avoid confusion about the status of this. ¶ Most of these templates are new to me. I wonder if all are useful. I looked at one example, Nihongo foot. About it, we're told "For more details on when to use this template, see Wikipedia:Manual of Style (Japan-related articles) and Wikipedia:Manual of Style (footnotes)"; however, the former doesn't seem to mention the string "Nihongo foot", while the latter one is about the citation of sources, which this is not. What it produces by default is what Wikipedia calls a reference; but actually it is not a reference in any normal understanding of the word. I appreciate that the group can be specified, effectively allowing the product of the template to be a non-reference footnote; but for this purpose its content is so short that I think most people would prefer to read this content inline rather than having to jump around the page to read it (particularly as a significant percentage of taps on a phone or tablet screen are slightly misplaced and then tend to have unexpected results). I'm aware that there has been an MfD about this template; but this was mostly a heated (though not venomous) argument between two people, one of whom has announced her intention to leave WP and the other of whom is indefinitely banned. I don't particularly want to see this template scrapped, but I'm also not keen that a multipurpose template should be usable for the same purpose and thereby encourage the proliferation of "references" such as these. -- Hoary (talk) 06:47, 20 February 2015 (UTC) strikethrough Hoary (talk) 07:09, 20 February 2015 (UTC)

I've only ever used {{Nihongo}}, but it definitely commonly uses English—for example, you might write: {{Nihongo|rickshaw|人力車|jinrikisya|"human-powered vehicle"}}—which outputs: rickshaw (人力車, jinrikisya, "human-powered vehicle"). If we're not going to keep all those pieces together in one template, then I don't see what prupose the template serves—I'd myself just start using {{lang|ja}} and {{transl|ja}}. Curly Turkey ¡gobble! 07:05, 20 February 2015 (UTC)

Yes of course. You're right, Curly Turkey. Clearly I need another coffee. (Oh, and I use nihongo2 rather than lang from force of habit.) -- Hoary (talk) 07:09, 20 February 2015 (UTC)

I only use nihongo out of force of habit myself—I've never tried the others. Curly Turkey ¡gobble! 07:13, 20 February 2015 (UTC)

About rickshaws: the example includes both a word of English ("rickshaw") and a phrase of English ("human-powered vehicle"). I am suggesting it is a bad idea to label either of these "English"; the template should make it easy to see how to include (for example) a normal English word for something, or a (non-English!) name by which something/body is known in English, and a translation gloss, using template parameter names carefully chosen (which I have not thought of yet). Equally, I think it is a bad idea to use the word "Romaji" (with or without the macron) to refer to transliteration, because it perpetuates the general confusion between characters and languages. Imaginatorium (talk) 07:20, 20 February 2015 (UTC)

Well, the gloss could be "|gloss=". Not so easy with the regular English, I guess ... "|term="? "|eng_term="? And for the rōmaji ... "|translit="? Or maybe "|transliteration=", as many people won't know what "translit" is short for (at least with the full term, you can look it up). Curly Turkey ¡gobble! 07:49, 20 February 2015 (UTC)

That's another advantage of moving the English out of the template – we wouldn't have to decide what to call it. I mean if we don't know what it is, why are we dealing with in the template? About the name for the transliteration, how about "hepburn", since it's a specific system and not one of the others? And about "lang|ja" and {{transl}}, I had never used either of those, so I looked. It seems that "lang|ja" is the absolute minimum. All it does is wrap the kanji in a "lang=ja" span, like Nihongo2. {{transl}} doesn't do much either. The main benefit seems to be that it displays "Japanese transliteration" in a tool tip when you mouse over it. That would probably be more useful for Arabic, where there are different systems, but for us not so much as we always use Hepburn. Compare that to Nihongo, which would still offer benefits even without the English, like putting the Japanese part in parentheses and italicizing the romanization. – Margin1522 (talk) 02:57, 21 February 2015 (UTC)

@Hoary:About Nihongo foot, my first instinct is to agree that we don't want to encourage these kind of footnotes. But looking at the TfD that you found, it mentioned Rozen Maiden as an example of where it's used, in game articles where the editors really want to include the Japanese names of the characters. Since the style of these articles is bit different, I'm starting to wonder whether we shouldn't just leave it up to their judgment. – Margin1522 (talk) 03:38, 21 February 2015 (UTC)

Ah, where for example quasi-German "Rozen Maiden: Wechseln Sie Weli Ab" is annotated as ローゼンメイデン ヴェヘゼルン ジー ヴェルト アップ Rōzen Meiden Vehezerun Jī Veruto Appu. Which certainly is bulky. Since there are quite a few of these notes, I'd have made a single table containing the same information. -- Hoary (talk) 03:47, 21 February 2015 (UTC)

Placement of the help question mark

Over at Talk:Yu Kanda#Little question mark @Siuenti: suggested that it may be more intuitive to place the question mark linking to Help:Installing Japanese character sets after the kanji/kana rather than after the Latin based romanji since the Japanese character set is the former, not the latter, and the question mark may suggest needing help with the romanji instead. Though Siuenti suggested this be done at the specific Kanda Yu article by substing the templates, I think it's better that this be discussed here and, should a consensus be in favor of the suggestion, implemented at the source across all articles rather than patching it through some individual articles (which, I imagine would create unnecessary discontinuity). As such, I'm moving the discussion for this here. Should the question mark be moved to the specific Japanese character set in question? ~Cheers, TenTonParasol 20:12, 14 May 2017 (UTC)

I didn't suggest it was better to do it just at that one article, I merely pointed out that it was possible when someone say it wasn't. Siuenti (씨유엔티) 20:28, 14 May 2017 (UTC)

Also I submit that this suggestion is plain common sense and should be carried out speedily and only discussed further if an actual objection emerges. Siuenti (씨유엔티) 20:30, 14 May 2017 (UTC)

 Done Okay, it's been changed. Good suggestion. ···日本穣 · 投稿 · Talk to Nihonjoe · Join WP Japan! 20:58, 14 May 2017 (UTC)

Remove the question mark?

(continued from above) The link to Help:Installing Japanese character sets was included back when native support for East Asian character sets weren't common, but with it being natively supported since Vista (2007), is it really needed anymore? —Farix (t | c) 21:14, 14 May 2017 (UTC)

Hmm it is on all major platforms do you think, including mobile, linux, stuff like that? I have no idea... who should we ask? Siuenti (씨유엔티) 21:20, 14 May 2017 (UTC)

Pretty sure displaying Japanese is built into every consumer OS shipped for a decade or more. Which includes all versions of Android and iOS, Windows since Vista, Mac OS since OS X. The only one I am not so certain of is Linux as I have not used it recently, but I can’t see it being that different – and Linux anyway is not a consumer OS, rather one for people who expect to do a bit of installing and configuring.--JohnBlackburne^words_deeds 21:31, 14 May 2017 (UTC)

I've decided to give this its own subsection. As for whom we ask, we ask the community. The question is, should we do a formal RfC or continue the informal discussion? —Farix (t | c) 21:39, 14 May 2017 (UTC)

Chinese characters itself doesn't seem to have the equivalent, so maybe not necessary after all. Siuenti (씨유엔티) 21:46, 14 May 2017 (UTC)

Yes, there is nothing equivalent in the main template used for Chinese, {{zh}}, and I can’t recall there ever being a demand for it. There is {{Contains Chinese text}} which seems a better solution, appearing only once, not interrupting the text, and more explicit and so clearer what it links to. But even that I think is falling out of use, or at least I see it less.--JohnBlackburne^words_deeds 22:47, 14 May 2017 (UTC)

I think a complete overhaul of this template would be a very good idea, but I think a narrow RfC, "Remove the question mark" would hugely simplify later discussion. This has been going on for ten years or more, and the last coherent justification I can find for needing "help installing Japanese character sets" (HIJCS) is in 2009: Template_talk:Nihongo/Archive_2#Remove_the_question_mark?. There was a comment somewhere (sorry, lost ref) about two years ago, but that referred to this Reddit discussion, and is about the minutiae of Japanese versus Chinese fonts. I do not think the current version of HIJCS helps with this at all.

What does HIJLS actually tell us? Mostly that "your operating system" will do this automatically; the nitty-gritty details are all about installing fonts in Unix variants, but I think this is because it is much easier to provide explicit command line instructions, whereas in Windows this would be pushing and poking around some menus. In practice, "out-of-the-box" Linux just works for everything this side of very obscure Indic character sets and similar. Arguably the most useful information is actually about template:nihongo, but using maximally confusing examples like {{nihongo|English|英語|eigo}}. Imaginatorium (talk) 08:13, 15 May 2017 (UTC)

Yes i agree, there is little point in confusing readers with a mark like that, when the far majority simply doesn't have the problem. I've been similarly annoyed by the Contains... templates for esp. Japanese and Chinese. These scripts have been supported by major operating systems for so long now, there just isn't a point to it any longer. —TheDJ (talk • contribs) 08:27, 15 May 2017 (UTC)

I can see removing it, but do we actually have evidence of it "confusing readers with a mark like that". I have never, even once, seen anything other than anecdotal evidence. ···日本穣 · 投稿 · Talk to Nihonjoe · Join WP Japan! 15:56, 15 May 2017 (UTC)

What sort of evidence would you need? A psychological study, involving a few hundred subjects? If something serves no real purpose, it is cluttering up the text, and removing it would be an improvement. OK, this is "anecdotal", but the other day I ran into one of these little question marks attached to a note on a reference: "Subscription required". When I hovered my mouse, I got a brilliant "explanation", which said "This site requires a paid subscription". I wasted my time; this is another example of a bit of probably once well-meant fluff, whose removal would improve things. Imaginatorium (talk) 14:16, 17 May 2017 (UTC)

Don't be so facetious. You're using an example from something else entirely as evidence that this one "wastes your time". That's not even a valid argument. ···日本穣 · 投稿 · Talk to Nihonjoe · Join WP Japan! 20:39, 22 June 2017 (UTC)

It’s confusing as it’s unclear. Does it take you to a page on the Japanese language? Or on the Japanese writing system? Or on the Romanisation used? Or on how to install Japanese on characters your system? The answer is the fourth of those, but that’s arguably the least obvious and least useful, with every consumer PC (Windows or Mac), phone and tablet (Android or iOS) sold today including full support for rendering Japanese. Of course users don’t complain about it, most will find the page they want with one or two more clicks. But in as much it leads many readers to a page they were not expecting it’s bad design.--JohnBlackburne^words_deeds 22:48, 17 May 2017 (UTC)

Hovering on the question mark tells you where it takes you, so it's not confusing at all. Your claim that it "leads many readers to a page they were not expecting" is completely unsubstantiated. ···日本穣 · 投稿 · Talk to Nihonjoe · Join WP Japan! 20:39, 22 June 2017 (UTC)

ping TheFarix, Siuenti, JohnBlackburne, Imaginatorium, Nihonjoe. I say we just get rid of it, but if people are uneasy, we could experiment by hiding this for a while and have the ban hammer come down on it at a later time. The following css would do this: —TheDJ (talk • contribs) 08:45, 22 June 2017 (UTC)

span[lang=ja] ~ sup.noprint {
	display:none !important;	
}

I'm fine with removing it, but let's not do it because it allegedly confuses people or links somewhere allegedly misleading because we don't have any evidence to suggest that. It was added, as Farix stated, back before Wikipedia adopted UTF-8 and back when Windows had either no or very crappy support for multiple languages. That is no longer the case, thankly, so I see no problem removing it because it is no longer really needed. ···日本穣 · 投稿 · Talk to Nihonjoe · Join WP Japan! 20:39, 22 June 2017 (UTC)

In fact, since I have seen no one argue in favor of keeping it, I've removed it. We now need to replace all instances of {{Nihongo4}} with {{Nihongo}} as the only difference is the lack of a question mark in the former. Since the latter no longer has the question mark, there is no need for the former. Can we agree on that so we can task a bot to the replacement? ···日本穣 · 投稿 · Talk to Nihonjoe · Join WP Japan! 20:46, 22 June 2017 (UTC)

It can just be redirected if no parameters have to be changed. Diff for current versions. PrimeHunter (talk) 16:07, 24 June 2017 (UTC)

 Done. ···日本穣 · 投稿 · Talk to Nihonjoe · Join WP Japan! 17:03, 24 June 2017 (UTC)

No comma between a word and its phonetic transcription

Please remove the comma between the Japanese word and its phonetic transcription. No commas appear between words and their phonetic transcriptions in phonetics literature. --Omnipaedista (talk) 11:42, 8 December 2017 (UTC)

Request: parameter to set a font-family to a specified value

I'd like to request a feature to enable setting a font-family parameter on the Template:Nihongo to specified content.
My motivation for asking this is that some characters don't display "correctly" with the default English Wikipedia font-family for the Desktop version, which is the Monotype Sans Serif.
Example: "a̠" and "ɯ̟". But if I <span>...</span> with the DejaVu Sans font-family, it gives: "a̠" and "ɯ̟"
So, I'd like the following example working with the IPA-ja parameter and with a customized font-family:
I did try setting <span>...</span> inside the Template:Nihongo, but it results in:
According to DES on Teahouse#Changing the font of a parameter value: "A template could accept a parameter indicating in what font its contents or part of them, would be displayed, but it would have to be coded into the template, it would not operate automatically. See {{tq}} where the |i= forces italics, for example."
[TRANSviada@talk ~]$ 19:53, 28 May 2018 (UTC)

You could simply add <span>...</span> inside the parameter to {{nihongo}}. Compare with and without span:

Shizuka Miura (三浦　静香, IPA: [ɕizɯᵝka̠ mʲiɯ̟ᵝɾa̠])

Shizuka Miura (三浦　静香, IPA: [ɕizɯᵝka̠ mʲiɯ̟ᵝɾa̠])

The IPA part changes to the other font while the other parameters are unaffected by the <span>...</span>. --Stefan2 (talk) 21:05, 28 May 2018 (UTC)

OwO thats simple and makes much sense. Many thanks Stefan2 for helping a template newbie 😅! [TRANSviada@talk ~]$ 22:15, 28 May 2018 (UTC)

Request: Add semicolon between Japanese text and Hepburn

I would like to request an edit to template editors to add a semicolon between the Japanese text and Hepburn romanization. The reason is in {{lang-zh}} and {{CJKV}}, their displays of the templates has semicolon to separate different aspects. —Wei4Green | 唯绿远大 (talk) 16:44, 13 November 2018 (UTC)

I think that a semicolon only is needed if the parameters need to be named, like "Kanji: 日本語; Hepburn: Nihongo" or "Shinjitai: 大日本帝国; Kyūjitai: 大日本帝國". If it's a simple "東京 Tōkyō" then I see no need for a semicolon. The other templates you listed seem to be designed for complex situations where parameters normally need to be named, like "Japanese: 竹島; Korean: 독도" or "simplified Chinese: 中国; traditional Chinese: 中國". --Stefan2 (talk) 21:35, 13 November 2018 (UTC)

@Stefan2: Yup, I agree. For example, currently in the article Asami Sato, the text in the curved brackets in "Asami Sato (Japanese: 佐藤麻美 Hepburn: Satō Asami) is a major character in..." has parameters named, but it doesn't have a semicolon separated, so it looks weird. —Wei4Green | 唯绿远大 (talk) 02:56, 14 November 2018 (UTC)

I think it would be better with a semicolon in the situation you mentioned. --Stefan2 (talk) 19:17, 14 November 2018 (UTC)

It would work still better with a comma and no link: "Asami Sato (Japanese: 佐藤麻美, Satō Asami) is a major character in..." Even "Tokyo (東京, Tōkyō)..." looks better, even for people who think the "Japanese" is needless in that case. — LlywelynII 04:24, 12 September 2019 (UTC)

I think in most cases for other languages a comma is used in such situations. This both looks better (IMO), and allows a semicolon to be used to separated the text/transliterations from other things like birth/death dates etc. I don't feel strongly about it though. —Ynhockey ^(Talk) 09:56, 4 December 2018 (UTC)

Yes, a comma should be used. It's been an issue raised at {{zh}} repeatedly but to no avail since the template's maintainer prefers to treat pinyin as a separate language in the highly-mistaken belief that the template is only be used in articles where Chinese will be the only language in the lede. It makes the template so malformatted that my default is to force everything into {{chinese}} infoboxes instead. His idea's far more accurate with Japanese articles, but it should still be using a comma.

That said, even a semicolon is better than nothing ({{chinese}} supports Japanese as well) and the real problem is wasting so much space and attention on the link to Hepburn in the first place. It's our only default romanization and should just be given after the characters in italics with no link at all: Bob (Japanese: 漢字, Bōbu). Another more elegant solution is used in Russian articles: use a small-text letter with a link for the curious: Bob (Japanese: 漢字, h Bōbu). — LlywelynII 04:22, 12 September 2019 (UTC)

I've noticed a related issue: MOS:NAME dictates that a semicolon, not a comma, should separate name pronunciation and dates of birth/death in biographical leads, which seems to the implementation on almost every non-Japanese biography I have checked. The closest that can be accomplished with this template is this hacky (and still incorrect due to the comma) example I found on Shinzō Abe's page, where IPA is part of the extra text: "Shinzō Abe (安倍 晋三, Abe Shinzō, IPA: [abe ɕin(d)zoː]; born 21 September 1954). Most Japanese biography pages don't have an IPA guide, and just put the dates in the extra field following the romaji, resulting in the also incorrect format of "romaji, born January 1, XXXX".

Something important to note is that the extra field is also commonly used for literal translations (both in leads and article bodies), for example mattress (敷き布団, shikibuton, lit. "spreading futon"), which is a situation where I think the comma works better than a semicolon would. Ideally this template would exclude the extra field entirely, but as for fixing things now, I'm not sure how it can be resolved without breaking things. Any thoughts on this? — Goszei (talk) 23:22, 29 October 2019 (UTC)

I also came here looking for guidance on this "semicolon, not a comma, should separate name pronunciation and dates of birth/death in biographical leads" issue. -Lopifalko (talk) 06:05, 24 January 2021 (UTC)

implementing nihongo and nihongo3 with module:lang

A recent somewhat related discussion elsewhere reminded me that {{nihongo}} and {{nihongo3}} do not render in the same way that {{lang}} and {{lang-ja}} render though it is common for the nihongo and lang templates to be used together in articles.

The obvious difference (to me) is that the nihongo templates lack tool tips. Because it was interesting for me to do it, I have hacked module:lang/utilities/sandbox, {{nihongo/sandbox}}, and {{nihongo3/sandbox}} to use _lang(), _lang_xx_inherit(), and _transl() from module:lang to render the Japanese and romaji parameters so that tooltips are available when readers mouse-over the rendered texts. You can see this in Template:nihongo/testcases and Template:nihongo3/testcases.

There are some cases in the test cases that are, to me, perplexing. Why, for example, would we ever want to use either of these templates to render only the English parameter value, or only the extra parameter value, or only the extra2 parameter value (or any combinations of only these three) without also rendering a Japanese or romaji value? It seems to me that one of {{{2}}} (Japanese) or {{{3}}} (romaji) should be required. Templates that don't include one of these should emit an error message and add the article to an appropriate tracking category so that the template can be fixed.

When |lead=yes and when {{{1}}} (English) has no value, why is it that {{nihongo}} does not prefix the romanization with the hepburn wikilink:

{{nihongo||日本|Nippon|lead=yes}} → Nippon (Japanese: 日本)

cf. when {{{1}}} (English) has a value

{{nihongo|Japan|日本|Nippon|lead=yes}} → Japan (Japanese: 日本, Hepburn: Nippon)

Why doesn't {{nihongo3}} support |lead=?

{{nihongo3|Japan|日本|Nippon|lead=yes}} → Nippon (日本, Japan)

I haven't touched {{nihongo2}} because that is more-or-less a call to {{lang|ja|...}} (further reason, I suppose, that {{nihongo}} and {{nihongo3}} should move to use module:lang).

Comments?

—Trappist the monk (talk) 14:49, 13 December 2019 (UTC)

And {{nihongo foot/sandbox}} and Template:nihongo foot/testcases. This live template supports the Help:Installing Japanese character sets link; the sandbox does not because none of {{nihongo}}, {{nihongo2}}, or {{nihongo3}} support that functionality.

Again, some of the supported 'functionality' makes no sense (only English or only extra or only extra2 or combinations of these).

The sandbox does not emit a footnote without at least on of Japanese, romaji, extra, or extra2; since these are the content of the footnote, if not are present it doesn't make sense to make a footnote. The sandbox fixes some text formatting errors that are present in the live template.

—Trappist the monk (talk) 20:00, 13 December 2019 (UTC)

There having been no comment, I have implemented the changes described above to {{nihongo}}. Errors detected by the template are categorized in Category:Nihongo template errors (0).

(edit) – {{nihongo3}} updated.

(edit) – {{nihongo foot}} updated.

Report problems here.

—Trappist the monk (talk) 18:57, 28 December 2019 (UTC) (nihongo) 14:32, 13 January 2020 (UTC) (nihongo3) 14:08, 18 January 2020 (UTC) (nihongo foot)

@Trappist the monk: This is another case of you making a major change that has sprinkled articles with error messages (I am working on the early Japanese emperors). It was only a few months ago that you got Wikipedia-wide backlash from doing the exact same thing on a larger scale. You are doing the right thing by taking it to the talk-page first, but these pages are not as well viewed. I urge you to take sweeping changes proposals to places such as WP:PUMP and/or do some testing beforehand. In my view marking parameters as "required" does not help if the outcome is not fixed. - Knowledgekid87 (talk) 15:06, 30 December 2019 (UTC)

Your ping did not work (you added it to a preexisting post).

In some form or another, the {{nihongo}} documentation has required the <kanji/kana> positional parameter ({{{2}}}):

12 October 2007

27 November 2009

4 April 2010

29 December 2019

The requirement has been present for more than a decade and in that decade+ editors have misused the template (despite the requirement) so that it does not do the things it is intended to do – a great example of that is the {{Administrative divisions of Japan}} template discussed below. The error messages are there to alert editors to problems with the template; when editors do not know, or realize, that a problem exists, they will not fix the problem.

Because I neither speak nor read Japanese, and because I cannot know the original editor's intent, it is not for me to hack a script to 'fix' these errors; that is a job best left to those who know and understand the purpose of the template in the article where it is used and the error message is displayed. Apparently, you do, so I have modified the template to show you the broken templates in articles that are of interest to you. Now you know that they are broken, whereas before you did not. With that knowledge, you can see to it that broken instances of {{nihongo}} in articles that matter to you, are repaired as you would have them repaired; not as some drive-by editor might 'fix' them.

—Trappist the monk (talk) 15:25, 31 December 2019 (UTC)

I'm not sure why, but it seems this change has made error messages appear on Template:Administrative divisions of Japan. Mccunicano☕️ 13:33, 31 December 2019 (UTC)

Yep, it did. Why? Because that template has a bunch of peculiar constructs like this:

{{resize|smaller|{{Nihongo|{{resize|larger|[[Subprefectures of Japan|Subprefectures]]}}}}}}

Not sure why the two {{resize}} templates; not sure why {{nihongo}} when there isn't any kanji/Kana or rōmaji text anywhere in that template. {{nihongo}} requires kanji/Kana and / or rōmaji so, neither of those being present, {{nihongo}} emits an error message to say that its requirements are not being fulfilled. It would seem to me that the above should be replaced with:

[[Subprefectures of Japan|Subprefectures]]

—Trappist the monk (talk) 14:07, 31 December 2019 (UTC)

Yeah it seems pretty redundant to use this template when there is no Japanese even used in those links. I'm unsure of why it was done in the first place. I'll go through and replace them. Mccunicano☕️ 16:10, 31 December 2019 (UTC)

If you're going to change this template, you ought to also change Template:Nihongo3 to match. They now have different formatting because this template marks the romanized text with lang="ja-Latn" (causing browsers to pick a Japanese font) and Nihongo3 doesn't. --Bigpeteb (talk) 21:41, 2 January 2020 (UTC)

Are you sure about that? lang="ja-Latn" should not cause browsers to pick a Japanese font because the Latn ietf tag identifies the text as using a Latin script, not a Japanese script. On my browser, Chrome 78.0.3904.108, this:

{{nihongo|English uses Latn font|Japanese uses Japanese font|Romanized uses Latn font}}

renders 'Japanese uses Japanese font' in a different font from that used for the English and romanized texts:

English uses Latn font<span style="font-weight: normal"> (<span title="Japanese-language text"><i lang="ja">Japanese uses Japanese font</i></span>, <span title="Hepburn transliteration"><i lang="ja-Latn">Romanized uses Latn font</i></span><span style="margin-left:.09em">)</span></span>

English uses Latn font (Japanese uses Japanese font, Romanized uses Latn font)

On my browser, that is exactly as it should be; is yours different?

I intend to update {{nihongo3}} and {{nihongo foot}} but, because no one said anything about these proposed updates, it was not clear to me what kind of torches and pitchforks uprising I would receive so I minimized my possible pain by limiting the update to {{nihongo}}. The other two should be updated together, assuming that, the holidays mostly over, there isn't a torches-and-pitchforks-uprising from returning editors.

—Trappist the monk (talk) 22:08, 2 January 2020 (UTC)

Yes, on latest Firefox (71.0), the lang="ja-Latn" text appears in a Japanese font, identical to the lang="ja" text. --Bigpeteb (talk) 00:09, 4 January 2020 (UTC)

Firefox has been to blame for similar issues with Arabic text. In this discussion, Template talk:Lang/Archive 8 § Arabic in western alphabet, one editor offered a couple of suggestions that can be used to override Firefox's font selection. Seems that someone should file a bug report with the Firefox people...

—Trappist the monk (talk) 00:40, 4 January 2020 (UTC)

Sometimes Japanese names have = signs in them, which appears to throw up this error even though it was presumably displaying properly before the change. I can fix it by doing this but is there a better solution? —Xezbeth (talk) 05:37, 6 January 2020 (UTC)

{{nihongo}} is not broken. This is a problem that occurs with any template that uses positional parameters as {{nihongo}} does.

When MediaWiki reads a raw template from an article's wikitext, positional parameters (those without names) are assigned 'numbered names'. In {{nihongo}}:

{{{1}}} gets the value <english>

{{{2}}} gets the value <japanese>

{{{3}}} gets the value <romaji>

When the value assigned to any of the numbered parameters has an assignment operator (=), MediaWiki interprets that as a named parameter.

So, for your example, Mediawiki sees a positional parameter {{{1}}} with a value of "Cock Robin Ondo" and a named parameter, クック that has the assigned value of ロビン音頭. MediaWiki then calls {{nihongo}} and passes it these two parameters. Nihongo does not recognize クック as a valid parameter name and did not get a <japanese> positional parameter so it emits the error message. (The previous version of this template simply ignored the |クック= parameter so it only displayed "Cock Robin Ondo".)

Your fix tells MediaWiki to ignore the assignment operator (content of the <nowiki>...</nowiki> tag is removed and replaced with a stripmarker which is, at the end, replaced by the content of the <nowiki>...</nowiki> tag). The common fix to these assignment-operator-in-positional-parameter-value problems it to use a number as a parameter name so while this does not work:

{{nihongo|"Cock Robin Ondo"|クック=ロビン音頭}} →

using |2=クック=ロビン音頭 does

{{nihongo|"Cock Robin Ondo"|2=クック=ロビン音頭}} → "Cock Robin Ondo" (クック=ロビン音頭)

—Trappist the monk (talk) 10:56, 6 January 2020 (UTC)

Not sure if you noticed, but there is also Template:Nihongo-s with 6 transclusions. Can probably be converted to this template, but haven't checked the templates. --Gonnym (talk) 16:02, 13 January 2020 (UTC)