Template talk:Infobox Chinese

Request for Literal meaning Japanese field

Following the example for Korean where there is a field lk for a literal meaning from Korean, a field such as lj would be useful and pertinent for literal meanings for the Japanese section. Waerloeg (talk) 08:26, 23 February 2022 (UTC)

Seconded toobigtokale (talk) 22:06, 15 August 2023 (UTC)

I assume this is still of interest, so I'm adding it to the to-do list. Remsense诉 00:35, 12 July 2024 (UTC)

Template-protected edit request on 30 April 2023

Add a "lang(x)_translit" field to compliment the lang(x)_content and lang(x) fields. Viewable via dropdown menu, like the languages already included in the template. Otherwise, add Kazakh and Kyrgyz fields to the template; I do not know why Tetum, a language spoken only on the island of Timor which has no significant Chinese community, is included in this infobox's fields, but not Kazakh or Kyrgyz which have their own autonomous areas in China. Yue🌙 07:06, 30 April 2023 (UTC)

This template is not just for China nowadays and really should be named something like Infobox transcription but that RM failed. I agree with adding Kazakh and Kyrgyz since they are (usually at least) not written in the Latin script. I'm not quite sure what exactly a translit field would do though. Usually edit requests are only for implementing edits in the sandbox rather than suggesting features but I won't close it quite yet since a few more qualified eyes may come here if it's open. --Trialpears (talk) 12:11, 30 April 2023 (UTC)

Closing per Trialpears' comment - it appears no qualified eyes are coming. * Pppery * _{it has begun...} 16:59, 6 May 2023 (UTC)

I assume this is still of interest, so I'm adding it to the to-do list. Remsense诉 00:36, 12 July 2024 (UTC)

Lang of bpmf parameter

I'm not sure where exactly it is implemented, but the value of the bpmf parameter ends up incorrectly marked up as zh-Latn instead of zh-Bopo. – MwGamera (talk) 12:03, 28 June 2023 (UTC)

I also ran into this. Remsense诉 22:07, 11 February 2024 (UTC)

Pe̍h-ōe-jī and Tâi-Lô

Pe̍h-ōe-jī has now a variant tag in BCP 47: Its text should emit nan-Latn-pehoeji instead of just nan-Latn. -- Error (talk) 12:40, 22 May 2024 (UTC)

Same about Tâi-uân Lô-má-jī Phing-im Hong-àn and nan-Latn-tailo --Error (talk) 12:49, 22 May 2024 (UTC)

Template-protected edit request on 11 May 2024

ENGVAR support: all I need is for labels to say "Romanisation" instead of "Romanization" (et al.) when an |engvar=b parameter is supplied, but I couldn't quite figure out how to do it in the corresponding module code. Remsense诉 08:07, 7 June 2024 (UTC)

@Pppery, you took a crack at this and then reverted? Izno (talk) 23:12, 1 August 2024 (UTC)

Yeah, I tried to implement it, and then I realized that more work than just my sandboxed edit would need to be done and lost the will to do it. I don't remember why I reacted to that realization by self-rollbacking, though. At this point I no longer care and anyone who wants to make another attempt at it is welcome to. * Pppery * _{it has begun...} 01:27, 2 August 2024 (UTC)

It seems clear that none of the people watching this queue are willing to code this up, so deactivating edit request. * Pppery * _{it has begun...} 20:58, 19 August 2024 (UTC)

Issue in dark mode or inappropriate use of infobox

It seems this infobox can sometimes have Chinese characters in the main image as in Kiwifruit and and sometimes photos as in Guanyin.

In dark mode (e.g. https://wiki.riteme.site/wiki/Guanyin?vectornightmode=1, not to be confused with gadget), the former works great as it is inverted, but the latter is broken. The invert doesn't appear to always work as in Hong_Kong.

Looks like the infobox description should be clarified to describe if photos are permitted and if so, be altered to support both use cases (only applying the invert when SVG text is given)

Thanks in advance for your help working on how to fix this! I am not sure how?

@User:Queen of Hearts fixed this like so but it would be good if there were clearer guidelines on this template about how to handle this situation! 🐸 Jdlrobson (talk) 18:53, 11 July 2024 (UTC)

The invert doesn't appear to always work as in Hong_Kong. You forgot to make the inversion apply to pic2 in addition to pic in Special:Diff/1227937682. I imagine introducing a separate parameter to InfoboxImage to explicitly allow inverting would be a safe way to handle it. – MwGamera (talk) 20:06, 11 July 2024 (UTC)

Not all SVGs are monochrome, so inversion won't fix them. Night mode is going to be a problem with any image with a transparent background (SVG or PNG), whether in an infobox or not.

It has long been common practice for the |pic= parameter of this template to contain an image of the object or person named if it can be depicted. Personally I think an image of characters is redundant. Kanguole 21:51, 11 July 2024 (UTC)

I see that this change, intended to make images of characters work with night mode, broke other images in night mode. It should be reverted, or at least enabled only if some extra flag is set. Kanguole 21:58, 11 July 2024 (UTC)

I know SVGs are cached as a raster render for display on pages, but surely there must be a sensible way to have currentColor (or unset) fills and strokes in SVGs render as a light tone in dark mode. Remsense诉 23:56, 11 July 2024 (UTC)

I guess the proper fix is to add a |picclass= parameter that can be set to skin-invert for black-on-transparent images of characters. Until this is fixed, a workaround is to use |pic2= for ordinary images and |pic= for the character images. Kanguole 14:45, 20 July 2024 (UTC)

I've reverted the edit. Izno (talk) 22:39, 1 August 2024 (UTC)

And followed it up with a marginally better workaround - it's always light mode for image backgrounds now. Izno (talk) 23:08, 1 August 2024 (UTC)

Just for some stats:

• About 60% of the uses of |pic= have a jpg
• 25% have an svg
• 10% have a png (I'd guess these are mostly transparent).
• And I'd guess the remainder of the 2k uses of |pic= have no image.

Probably what should be done is not actually a picclass but instead the suggested workaround, or even to embed that workaround in the template-proper as |character_image= and |other_image= or some such. Izno (talk) 22:34, 1 August 2024 (UTC)

Missing language name when hide is "no"

Title
Chinese	[[[wikt:N/A|N/A]]] Error: {{Lang}}: Latn text/non-Latn script subtag mismatch (help)
Romanization	This field is wuu, but you wouldn’t know that.

Northern Moonlight 08:19, 2 September 2024 (UTC)

A poor choice of parameter-name / parameter-value pair that was in the wikitext version of this template when it was converted from wikitext to Module:Infobox multi-lingual name. The last wikitext version of this template calls {{Infobox Chinese/Chinese}} which had then (and still has today), this for Wu:

| header20 = {{#if:{{{wuu|}}}{{{lmz|}}}{{{suz<includeonly>|</includeonly>}}} | {{#ifeq: {{{hide|}}} | no | | [[Wu Chinese|Wu]] }} }}

Because of that poor choice made so long ago, when I converted the mess of wikitext templates that supported {{Infobox Chinese}} to Module:Infobox multi-lingual name I left a comment in the Lua code:

local show = 'no' ~= args.hide or nil; -- make boolean-ish for controlling display of headers; |hide=no means show transcriptions without collapsed header

If we are to believe these search results there are about 1000 articles that have |hide=no so, were we to change to support both |hide=no and |hide-header=yes (or some such), I imagine that an ambitious editor(s) could make the necessary replacements in a reasonably short period of time after which |hide=no goes away forever.

But, because life is never simple, there are several subtemplates of {{Infobox Chinese}} that also support |hide=. To be consistent with {{Infobox Chinese}}, those, and the articles that call them, would also need to be updated.

—Trappist the monk (talk) 14:20, 2 September 2024 (UTC)

Image of characters

In many iterations of {{Infobox Chinese}}, people are including an image of the Chinese characters for the word, as well as the word itself in the infobox. I feel like this is redundant and takes up space; you're seeing the same word printed twice. Should this practice be discouraged? seefooddiet (talk) 07:03, 28 November 2024 (UTC)

Example: Chopsticks. I know the image has a caption and a footnote; that info could possibly be moved to appear in a footnote after the text in the alternate Chinese name parameter or just be explained in the article body somewhere. seefooddiet (talk) 07:04, 28 November 2024 (UTC)

According to User:White whirlwind, from WT:ZH#Provincial infoboxes: The SVGs had dual rationales: One, I thought them aesthetically pleasing; two, they fix the problem of the huge discrepancies in how Chinese fonts are displayed across different browsers, operating systems, and devices. In all likelihood, they are best used only on articles whose titles are Chinese names or words. Kanguole is correct that he objected to them (here, I think) back when I first added them, but nobody joined him.

I still feel as if we should probably not have them in most of the places we do, but I don't want to rock the boat in case there's aspects I'm not considering. Remsense ‥ 论 07:30, 28 November 2024 (UTC)

Thanks for looking into this. No opinion on aesthetically pleasing; purely interested in practical value to readers. I could see some merits to the points about discrepancies in device displaying, but I'm still not thoroughly convinced.

To your suggestion about not having them, I think we should spell out when to use these images in one or more highly visible places (template docs, maybe in the visual editor when populating the template). Otherwise I can envision well-meaning users trying to put these images in in basically every article by default out of desire for consistency. seefooddiet (talk) 07:40, 28 November 2024 (UTC)

My objection was that images of characters were redundant clutter, since the {{Infobox Chinese}} already has the characters. Now that the technical issues are much weaker, the only remaining argument for them is as decoration. Kanguole 08:45, 28 November 2024 (UTC)

I suspect that this practice arose from a common technical issue that existed in the early 2000s; displaying Chinese characters on English-language OSes at the time required manual installation of a Chinese font.

Today, Chinese characters that are in the ranges of U+4E00–U+9FA5 and U+3400–U+4DB5 get good font coverage; they do not need an image. 172.56.232.194 (talk) 07:25, 28 November 2024 (UTC)

I wouldn't go quite that far. Most people are browsing on their phones etc. I can easily see that if one cannot read the characters, they also cannot identify them at body-text size. If we care about the visual forms of the names in certain articles (which, to a degree we do, even though Wikipedia is not a dictionary) this makes a certain amount of sense even if the ordinary text renders just fine. Remsense ‥ 论 07:32, 28 November 2024 (UTC)

I think there's possibly value to prominently displaying the characters. However, I'd prefer alternatives where we don't display the characters twice. Maybe we could adjust the infobox body text larger? Readability for Chinese characters especially suffers from small font size. seefooddiet (talk) 07:44, 28 November 2024 (UTC)

I think there's value in the "traditional" regular script presentation, since many of the articles where we'd want to do this existed before the late 20th century, and as such it would feel a little silly seeing 荀子 emblazoned at the top of a page in 48pt Gothic. Remsense ‥ 论 07:56, 28 November 2024 (UTC)

Could you reword? What is "traditional" regular script presentation? seefooddiet (talk) 08:02, 28 November 2024 (UTC)

The style the images are in—one reflecting handwritten or premodern printed regular script. Kai (楷) fonts, as it were. Remsense ‥ 论 08:05, 28 November 2024 (UTC)

Ah I see. Hm, will think seefooddiet (talk) 08:07, 28 November 2024 (UTC)

Enclosing Chinese characters inside {{serif}} produces the more old-school 宋體 style, which is very slightly less legible at small type sizes. I don't think we should employ this in infoboxes on grounds of anachronism, because more anachronisms lie just beyond. Folly Mox (talk) 10:29, 6 December 2024 (UTC)

Unless we just want large-point serif glyphs instead of svgs. Folly Mox (talk) 10:31, 6 December 2024 (UTC)

This was part of the "aesthetic" motivation I mentioned. There aren't many other places to, for example, include a seal script image and have it still look encyclopedic and orderly.  White Whirlwind  15:59, 28 November 2024 (UTC)

Seal script I see potential value for, although I could see arguments that they better fit in the body instead. For cases where the image is close in appearance to standard Wikipedia fonts I'd be skeptical. seefooddiet (talk) 16:40, 28 November 2024 (UTC)

Characters might be fine, what isn't fine is an image of a character. That can't be read by screen readers or copied by selecting them. If a specific font and size is wanted, just wrap the text in a span and use define them. Gonnym (talk) 07:38, 28 November 2024 (UTC)

I wish Wikipedia actually displayed SVGs as SVGs, because this wouldn't be a problem. There's a way we can do that, but it's some sort of hack and people would start screaming. Remsense ‥ 论 07:57, 28 November 2024 (UTC)

Conversation hit a lull. My interpretation of things so far is that there's no strong consensus on what to do with the infobox images. They probably shouldn't be universal; some scenarios they're probably unhelpful and add to bloat. Showing calligraphy or seal script in the infobox may be useful, although it's not straightforward. Is this a fair assessment of our discussion? seefooddiet (talk) 05:53, 1 December 2024 (UTC)

Having just turned up after somehow missing all posts to Wikipedia talk:WikiProject China for like a month? Even though I have it watchlisted? I think that images of calligraphy and pre-楷書 scripts have value in infoboxes given the subject is an appropriate fit.

Chopsticks is a fine example of where these are not necessary.

I think body text is generally a worse placement for any early script or calligraphic script images we may want to retain, with some exceptions, like articles about particular script forms or styles. Keeping the most relevant Chinese characters consolidated within the infobox makes more sense than scattering superficially related glyph images in arbitrary body subheadings.

Chinese text display has indeed progressed a great deal in the past twelve or fourteen years: no need to teach your device how to decode both Big5 and GBwhateveritwas; Unicode is ubiquitous and everything supports it.

As a mobile-only editor / reader, I can confirm that modern screens (anything since maybe 2012?) display Chinese characters legibly at standard type sizes, including "small". Confirming also that mobile users understand how to pinch zoom for display elements requiring magnification.

I think it would be appropriate to do a broad sweep of articles transcluding this infobox template, with the goal of removing images where they add no value over standard unicode character display, but I don't think sunsetting support for the image parameter is called for.

(Unless it's actually less work to add a new "alternate script image" parameter for articles where early or calligraphic scripts are appropriate, migrate those manually, and deprecate the current parameter name. It would be helpful here to have some data on the scope.) Folly Mox (talk) 10:52, 6 December 2024 (UTC) dumb idea struck 11:39, 6 December 2024 (UTC)

A search for hastemplate:"Infobox Chinese" insource:pic returns 2461 articles; a search for hastemplate:"Infobox Chinese" insource:piccap (to avoid matching pic in other templates / prose) returns 1950 articles.

Special characters like = and | are treated as "greyspace" characters, and the standard search will not differentiate them from each other or from whitespace, which makes searching for particular parameters difficult. More precise results could be obtained by composing the hastemplate filter with an applicable regex, but it might time out due to regex processing overhead.

No idea how many of these cases are images of characters set in forms not duplicative of unicode supported glyphs. Folly Mox (talk) 11:13, 6 December 2024 (UTC)

"seal script" 153; "seal.svg" 61; 篆 ("seal") 9; remainder of ideas returned mostly false positives. Not sure there's any way to find all articles displaying Commons images across a whole category, like c:Category:Seal script or c:Category:Chinese calligraphy. Folly Mox (talk) 11:24, 6 December 2024 (UTC)

See the figures from Izno in #Issue in dark mode or inappropriate use of infobox above. Kanguole 11:26, 6 December 2024 (UTC)

Foiled again by Minerva not autoexpanding existing subheadings! Ok, the scope seems less than my naive search predicted; advancing out to later results does in fact show many articles with empty |pic= params, and also plenty where |pic= holds an image that isn't even Chinese characters (so deprecating the parameter would be a bad idea; struck). Folly Mox (talk) 11:39, 6 December 2024 (UTC)

Edit request 29 December 2024

Description of suggested change: Take out "lmz" Shanghainese long-short romanization, which seems to only be used on WP + associated article was deleted six years ago. (Alternatively, we could put in another tag for a generalized "Shanghainese romanization" tag, perhaps we should change the code as lmz is used for a completely different language by ISO.) Replace with "wgn" Wugniu, the most commonly used romanization for Wu (meant to be used across varieties) and the one used on Wiktionary. MSG17 (talk) 20:23, 29 December 2024 (UTC)

Diff: See here.

-MSG17 (talk) 20:23, 29 December 2024 (UTC)

Also added a param for Wu IPA, which may not be as helpful, but I've heard that in the absence of a standard romanization IPA is pretty common. MSG17 (talk) 20:27, 29 December 2024 (UTC)

Which Wu is that? Shanghai, Suzhou, Wenzhou or some other? They're not the same. Kanguole 23:19, 29 December 2024 (UTC)

Wugniu is meant to be used by many varieties of Wu. You can see here for various lects that Wugniu can be used to represent. wikt:User:ND381/Wugniu contains an incomplete chart comparing the marking of tones and orthographic differences of Wugniu across each Wu language (though I think Wiktionary mainly focuses on Taihu/Northern Wu). MSG17 (talk) 16:19, 30 December 2024 (UTC)

Right, my comment was more directed at IPA, for which the variety needs to be specified. Kanguole 16:26, 30 December 2024 (UTC)

Well, I would think it might be possible to just add a label with the variety and maybe use newline/unbulleted lists for multiple ones. MSG17 (talk) 20:38, 30 December 2024 (UTC)