Template talk:Transliteration

Articles containing x-language text

I may be mistaken, but unlike the Lang template, it doesn't look like the Transliteration template adds the category "Articles containing x-language text" to articles using it. Some articles contain only transliterated foreign-language text, not the original script, and won't have the category applied to them. Why are the two templates different in this regard? flod logic (talk) 08:32, 18 January 2023 (UTC)

For my part I can only say that because the wikitext version of this template did not categorize by language, the Module:lang version of this template does not categorize by language. You might ask Editor Dbachmann why the original wikitext version did not do such categorization.

—Trappist the monk (talk) 14:11, 18 January 2023 (UTC)

Thanks for your reply. I don't know enough about the technical side of it, but as a frequent language tagger, it would make sense to me to have them both categorize by language for consistency's sake. Is it possible to change that? flod logic (talk) 13:51, 21 January 2023 (UTC)

Because {{transl}} accepts both language and script tags, it seems to me that this template should not share the categories:

• Articles containing <language name>-language text
• Articles with text in <language name>
• Articles containing explicitly cited <language name>-language text

So, if we do this, perhaps these category names:

• Articles containing <language name>-language transliteration
• Articles with transliteration from <language name>
• Articles containing explicitly cited <language name>-language transliteration
• Articles containing <script name>-script transliteration

And there is this: Creating these categories in Module:Lang will all-of-a-sudden create links to about a thousand redlinked categories that will need to be created. That can likely be automated if we create a template that can add appropriate text to the transliteration categories much like {{Non-English-language text category}} does for the language categories.

Not a simple task. Worth doing?

—Trappist the monk (talk) 20:01, 22 January 2023 (UTC)

Private-use language tags

Can the Transliteration template be modified to support Private-use language tags? I have been expanding Wikipedia's coverage of ancient history, and I am finding myself needing for the transliteration template to render Private-use language tags, but it seems that it does not currently do so.

Seeing as the expansion of Wikipedia's ancient history coverage would inevitably make it a necessity, can I request for the Transliteration template to be modified so that it can render Private-use language tags? Antiquistik (talk) 12:01, 19 January 2023 (UTC)

Source

Hi @Jonesey95, why did you revert me?[1] “Source text to be transliterated,” foreign-script text like барахло, is never to be entered into this template. Only its corresponding Latin-alphabet target transliteration, like barakhlo.  —Michael Z. 02:05, 6 June 2023 (UTC)

My mistake. I was suspicious of the edit, because the documentation had been stable for a while, and I misread the documentation. – Jonesey95 (talk) 03:07, 6 June 2023 (UTC)

Thank you.  —Michael Z. 12:34, 6 June 2023 (UTC)

Edit request 15 January 2024

Description of suggested change: Hi. I want to change the Proto-cuneiform listing to point to the main article ie "Proto-cuneiform" from the current "Proto-cuneiform numerals", which is a sub-article. Thanks.Ploversegg (talk) 02:44, 15 January 2024 (UTC) Diff:

−

ORIGINAL_TEXT

+

CHANGED_TEXT

Ploversegg (talk) 02:44, 15 January 2024 (UTC)

Not done. You have not provided an explicit 'change-this-thing-in-this-part-of-the-template-to-this-other-thing' description. Don't make us guess at exactly what you want us to do.

—Trappist the monk (talk) 03:56, 15 January 2024 (UTC)

Ok, let me stare at it and figure out the right words. This is my first time working with this template.Ploversegg (talk) 04:00, 15 January 2024 (UTC)

Description of suggested change: Lets see if I can do better this time. I would like to request that in "Template:ISO 15924 script codes and related Unicode data" in the line "Pcun" that "Proto-cuneiform numerals" be changed to "Proto-cuneiform" corresponding to the article Proto-cuneiform. Thanks.Ploversegg (talk) 19:53, 15 January 2024 (UTC)

Diff:

−

ORIGINAL_TEXT

+

CHANGED_TEXT

Ploversegg (talk) 19:53, 15 January 2024 (UTC)

{{ISO 15924 script codes and related Unicode data}} is not part of {{transliteration}} per se. It is its own template. It calls {{ISO 15924 script codes and related Unicode data/row}} which has this:

[[{{ISO 15924/wp-article|1={{{alpha4|}}}}}|{{ISO 15924 name|1={{{alpha4|}}}}}]]

where {{{alpha4|}}} is the ISO 15924 tag (in this case pcun). {{ISO 15924/wp-article}} defines pcun as Proto-cuneiform numerals but {{ISO 15924 name}} defines pcun as Proto-Cuneiform.

The best place to post this edit request is at the template where the change will be made. Or, because you have extended confirmed editing rights, you can fix {{ISO 15924/wp-article}} yourself.

—Trappist the monk (talk) 20:31, 15 January 2024 (UTC)

Thanks. Now my brain hurts. I'm going to try to edit Template:ISO 15924/wp-article and will hopefully not break Wikipedia in the process. Amazing how one can edit for many years and still not be aware of all the stuff under the hood.Ploversegg (talk) 20:54, 15 January 2024 (UTC)

Template error

The inclusion of the equal sign (=) in texts results in an error message instead of the text being displayed. This causes serious issues for the transliteration of languages like Hurrian, Urartian, Ancient Egyptian, and Luwian, whose transliteration requires the use of the equal sign.

For example:

• Hurrian "pašš-ēt-i=t=ān," when put through the template as [undefined] Error: {{Transliteration}}: no text (help), results in an error;
• Urartian "šidišt=u=nə," when put through the template as [undefined] Error: {{Transliteration}}: no text (help), results in an error;
• Ancient Egyptian "Ꜥnt Ꜥstrt n=f m jkm," when put through the template as [undefined] Error: {{Transliteration}}: no text (help), results in an error;
• Luwian "a=wa=mu zan allantallin ammis nannis piyatta," when put through the template as [undefined] Error: {{Transliteration}}: no text (help), results in an error.

Can this issue be fixed? Antiquistik (talk) 14:02, 23 March 2024 (UTC)

When an equal sign exists in an unnamed positional (or unnumbered) template parameter, MediaWiki interprets the positional parameter value as a 'named' parameter/value pair. This is not exclusive to {{transl}} and has been ever thus for all templates that use positional parameters. So, in |pašš-ēt-i=t=ān, pašš-ēt-i is interpreted as the parameter name and t=ān is that parameter's value. Because pašš-ēt-i is not a parameter name that {{transl}} recognizes, it is ignored. The no text error message occurs because {{transl}} did not get a valid second (text) parameter. The commonly used work-around for this is to number the second positional parameter:

{{transl|xhu|2=pašš-ēt-i=t=ān}} → pašš-ēt-i=t=ān

—Trappist the monk (talk) 14:47, 23 March 2024 (UTC)

Thanks! Could you add this information to the main template page so other users won't experience similar issues in the future? Antiquistik (talk) 14:57, 23 March 2024 (UTC)

Requested move 5 April 2024

The following is a closed discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. Editors desiring to contest the closing decision should consider a move review after discussing it on the closer's talk page. No further edits should be made to this discussion.

The result of the move request was: not moved. Withdrawn; primary given reason for a preference mooted by ensuring AWB doesn't autoreplace transl with transliteration. Thanks, everyone! (non-admin closure) Remsense诉 06:54, 7 April 2024 (UTC)

---

Template:Transliteration → Template:Translit – The extra eration really contributes to a readability issue when there are lists, tables, or any preponderance of this template in an article—I get that {{transl}} is undesirably ambiguous, but frankly: the shorter the better. I would even prefer {{tlit}} if other people will let me get away with it.

No one in the 2022 move discussion mentioned an explicit issue with {{translit}}, so hopefully folks would be okay with it now. Remsense诉 18:25, 5 April 2024 (UTC)

• Oppose per the explicit issue I mentioned in the 2022 discussion: "Template function should be clear from the template name" (this is a quote from WP:TMPG, a guideline). "Transliteration" is the proper canonical name for this template. You are free to use the redirect, and other editors should not replace it in the wikitext, per WP:NOTBROKEN, another guideline. I do not object to the creation of a redirect at {{tlit}}. – Jonesey95 (talk) 20:30, 5 April 2024 (UTC)

  That's not the case, unfortunately, as WP:AWB does replace {{transl}} et al with {{transliteration}}, roughly in line with other substitutions, which I agree with in principle because consistency between articles is nice. I did specifically poke around the tool talk page asking whether such an replacement could be removed, but no one seemed interested, so I guess I'm just realizing this is a bit of forum shopping to that effect, whoops.

  I suppose the function of "translit" would be clear to me, and perhaps to most that would be in the business of using such a template, especially in context. Remsense诉 00:51, 6 April 2024 (UTC)

  Oh, and to explicate: I think it's fair to say that "translit" is at least a hair less clear for editors, but when an article calls it 300 times, that's an extra 2.1kB on the article. Combined with a lot of the structural load in many template-heavy articles, extra length often makes articles physically more difficult to edit without going section by section. I don't want to be contrarian, but it feels like there are more concrete reasons to consider this move, and the case that {{transliteration}} is better solely for reasons of clarity is largely theoretical, as I haven't seen anyone say its meaning is actually unclear or confusing I hope you see how the flexibility in that guideline allows for us to disagree on this point. Remsense诉 01:13, 6 April 2024 (UTC)

  I do not think that there is consensus for AWB to rewrite "translit" or "transliterate" to "transliterate". I would support removal of those two particular redirects from Wikipedia:AutoWikiBrowser/Template redirects. As the top of that page says, Before adding a rule here, you must ensure that there is consensus in favour of the template renaming. Those two redirects were added by Mclay1 in February 2023; that editor might be able to link to a relevant discussion. – Jonesey95 (talk) 14:03, 6 April 2024 (UTC)

  Despite that message, consensus is rarely if ever established beforehand for individual templates. Previous discussions have established consensus for bypassing template redirects unless there is a reason not to. If there is an objection, as in this case, the template can easily be removed from the list. I don't have a problem with that. MClay1 (talk) 06:48, 7 April 2024 (UTC)

  In that case, I will be doing so and withdrawing this move request as my main reason for preferring it is moot. Thank you for the engagement @Jonesey95, @Mclay1 et al. Remsense诉 06:52, 7 April 2024 (UTC)

• support 'Translit' is clear an unambiguous, also lang-xx templates use 'translit' and it's annoying to type out the whole thing when switching away from them—blindlynx 00:36, 6 April 2024 (UTC)
• Oppose per the WP:TMPG. The current name describes the template's function clearer than the proposed name. Redirect from the shorter name exists, and if the issue is a few characters saved in the database, then talk to the folks at WP:AWB. -- Netoholic @ 13:01, 6 April 2024 (UTC)
• Oppose – Proper English names for templates are much easier to understand. There is nothing stopping editors from using shortcuts in articles if they desire. MClay1 (talk) 06:49, 7 April 2024 (UTC)
• Oppose per WP:TMPG. Nardog (talk) 06:52, 7 April 2024 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Style guidelines for multiple templates

See also: Template talk:Translation § Style guidelines for multiple templates

Are there style guidelines for the use of both {{translation}} and {{transliteration}} in conjunction with {{lang}}? E.g., "להד״​מ​ (lahada"m), לא ה​י​ו ד​ב​ר​י​ם מ​ע​ו​ל​ם (lo hayu dvarim meolam) - transl. such things never were, no way" -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 13:46, 14 April 2024 (UTC)

Underdocumented error type

(Redirected from Category talk:Transliteration template errors)

In the middle of an unrelated Citation bot cleanup run, I found myself at Scythian languages, where {{transl}} is emitting oodles of errors of the type transliteration text not Latin script. This error is not documented at the help link, which points to Category:Transliteration template errors.

It's pretty clear what this means, and I tried to fix by subbing in {{lang}} per the documentation here, but that broke several links where the {{transl}} output was piped to an internal link. (Upon review, this is already broken.)

Not sure if that ever worked or how to fix it, or what the behaviour of {{transl}} used to be for reconstructed languages, when fed mostly Latin script with a few pronunciation glyphs like ϑ, δ, and γ. I'm not comfortable updating Category:Transliteration template errors to address the error mentioned, in case different use cases call for different fixes.

Dimly aware of the recent Module:Lang-related overhauls, Folly Mox (talk) 15:46, 30 November 2024 (UTC)

Error message documented; copyediting probably desirable because I suck at documentation – it is known.

This construct (and others like it at Scythian languages):

[[Ariapeithes|{{transl|xsc|*Ariyapaiϑah}}]]

violates the first sentence of the {{transliteration}} documentation:

This template is used to mark up text transliterated or romanised from a non-Latin alphabet script to Latin alphabet script.

In *Ariyapaiϑah, ϑ is U+03D1: GREEK THETA SYMBOL; not a Latn-script character.

To avoid the error message, one might write:

[[Ariapeithes|{{lang|xsc|*Ariyapaiϑah|nocat=yes|italic=yes}}]] → *Ariyapaiϑah

{{transliteration}} knows that Unicode does not have a Latn-script theta (θ; U+03B8: GREEK SMALL LETTER THETA). See Template talk:Lang § Non-latn text/Latn script subtag mismatch errors in ancient Iranian articles. I gotta wonder if the decision to accept θ was correct. I guess I want to see that θ really is needed for romanization. If it is, and a sufficient argument can be made for ϑ, perhaps we can collect a carefully curated list of other non-Latn-script characters that may be accepted as 'Latn'.

—Trappist the monk (talk) 17:41, 30 November 2024 (UTC)

It appears something changed recently? -- I am here from the Etruscan_language#Syllabic_theory page/section where the is a new(?) transliteration text not Latin script error. I could not find a clear definition from the help pages of what `Latn` script was. It seems to include all sorts of non Latin character modifiers, and glottal stops for Arabic, and θ (theta) randomly, as also required for Etruscan, but not φ (phi), which is equally required. This seems like many transliteration schemes will have special 'extra' symbols affected. Etruscan and Scythian are two examples. Assuming transliteration schemes must be Latin or 'Latn' (without being clear what that means, or why) seems an unnecessarily limiting restriction. I reverted another editor's good faith edit which added a IPA phi, (IPA is Latn?) which presumably removed the error message before I understood the reason behind it. I'm looking for a correct way to correct this and remove the error message, but am not sold on the seemingly arbitrary character restriction for un-specified transliteration schemes. Salpynx (talk) 23:15, 2 December 2024 (UTC)

Hey @Salpynx, it was my edit that you reverted. I am also baffled by the restrictiveness of transliteration templates.
For Greek "γ χ φ" - "ɣ ꭓ ɸ" are considered Latn by the template (and look almost the same as the Greek characters in Wikipedia's fonts), Greek θ is mostly accepted, however Sigma "σ" doesn't seem to have any accepted equivalent. I've tried to remove the errors as best I could; you're welcome to revert the edits again if it is imperative for the Etruscan transliterations to use Greek characters, however unless the templates are changed swiftly the errors will stay. Samsattet001 (talk) 12:41, 6 December 2024 (UTC)

 Comment: Came here to check what's happening. Eurovision: Europe Shine a Light has a lot of errors. — IмSтevan ^talk 23:48, 8 December 2024 (UTC)

By writing {{transliteration|Ethi|ፍቅር ልቤ}}, you are saying that ፍቅር ልቤ is a Latn transliteration of Ethi (Ethiopic-script) text. Clearly that is not correct. You would be better served were you to write:

{{lang|am|ፍቅር ልቤ}} → ፍቅር ልቤ

or

{{langx|am|ፍቅር ልቤ}} → Amharic: ፍቅር ልቤ

Same applies for the Belarusian and Ukrainian items in that list.

—Trappist the monk (talk) 00:24, 9 December 2024 (UTC)

I believe that the template should be changed so that it is not exclusive to Latin transliteration. 2A02:FE1:9293:F00:D1E5:C603:2AD2:4855 (talk) 13:59, 13 December 2024 (UTC)

This also occurs with ъ and ь in articles about Old Church Slavonic. 2A02:FE1:9293:F00:AC7C:5412:BC9B:226A (talk) 18:42, 22 December 2024 (UTC)

Please allow modifier letters widely used in transliteration

ALA-LC and other romanization schemes use some of the following modifier letters:

• ʼ (U+02BC MODIFIER LETTER APOSTROPHE)
• ʻ (U+02BB MODIFIER LETTER TURNED COMMA)
• ʿ (U+02BF MODIFIER LETTER LEFT HALF RING)
• ʾ (U+02BE MODIFIER LETTER RIGHT HALF RING)
• ʺ (U+02BA MODIFIER LETTER DOUBLE PRIME)
• ʹ (U+02B9 MODIFIER LETTER PRIME)

Locoluis (talk) 19:21, 9 December 2024 (UTC)

They are allowed:

• {{transliteration|und|ʼ (U+02BC MODIFIER LETTER APOSTROPHE)}} → ʼ (U+02BC MODIFIER LETTER APOSTROPHE)
• {{transliteration|und|ʻ (U+02BB MODIFIER LETTER TURNED COMMA)}} → ʻ (U+02BB MODIFIER LETTER TURNED COMMA)
• {{transliteration|und|ʿ (U+02BF MODIFIER LETTER LEFT HALF RING)}} → ʿ (U+02BF MODIFIER LETTER LEFT HALF RING)
• {{transliteration|und|ʾ (U+02BE MODIFIER LETTER RIGHT HALF RING)}} → ʾ (U+02BE MODIFIER LETTER RIGHT HALF RING)
• {{transliteration|und|ʺ (U+02BA MODIFIER LETTER DOUBLE PRIME)}} → ʺ (U+02BA MODIFIER LETTER DOUBLE PRIME)
• {{transliteration|und|ʹ (U+02B9 MODIFIER LETTER PRIME)}} → ʹ (U+02B9 MODIFIER LETTER PRIME)

—Trappist the monk (talk) 23:19, 9 December 2024 (UTC)

Right half ring is not allowed for Arabic: {{transliteration|ar|ʾ}} gives ʾ. I see this error at Sabians#Etymology. I've noticed similar problems with {{langx}} and Greek γ in transliterations of Old Turkic; presumably the root cause is the same. Why is the template so hardass about this? Such behaviour violates Postel's prescription and makes contributing to the encyclopedia a pain. Hairy Dude (talk) 10:22, 13 December 2024 (UTC)

Right half ring is a 'modifier letter'. That naming suggests that it 'modifies' something. In isolation, as it is here:

{{transliteration|ar|ṣ}}-{{transliteration|ar|b}}-{{transliteration|ar|ʾ}}

right half ring modifies nothing. When there is something for right half ring to modify, as in:

{{transliteration|ar|Ṣābiʾ}} → Ṣābiʾ

then right half ring is accepted. Even when it doesn't actually modify anything in the text that includes it, as I demonstrated above, right half ring is accepted.

This same is true for all of the other modifier letters listed above: in isolation, there is nothing to modify so it is meaningless, and perhaps even misleading, to individually markup these modifiers as romanizations.

If there is a transliteration/romanization standard for Old Turkic, en.wiki does not, apparently, have an article describing it; see this search. I have to wonder then if the use of Greek gamma in Old Turkic romanizations is something someone made up which other editors then parroted. I suspect that the common misapplication of the Cyrillic small o-with-combining-macron in the Japanese romanization: 'Tettei Kо̄sen' is the same sort of parroting. There is a Latin small-o-with-macron: 'ō' just as there is a Latin small-gamma 'ɣ'.

—Trappist the monk (talk) 16:06, 13 December 2024 (UTC)

"in isolation, there is nothing to modify" -- but in statements on orthography, some characters are specifically meant to be shown in isolation.
`◌` should be an accepted placeholder character, since it is necessary for displaying transliterated graphemes that are not independent letters due to a transliterative convention, but do represent a distinct phoneme, and not just a modification of another one.

example: in the table in Nahuatl#Writing, in the last row, on character `'`
- it's a glottal stop that just happens to be represented as a diacritic that can be put on whichever vowel happens to be at a given position in a word.

--Oazrin (talk) 07:14, 28 December 2024 (UTC)

U+25CC: DOTTED CIRCLE is accepted when accompanied by one or more Latin-script characters:

{{transliteration|en|a◌z}} → a◌z

When written by itself, there is no need to wrap it in the {{transliteration}} template because the code point is just a code point. This same is true for U+02BC: MODIFIER LETTER APOSTROPHE (last row of table in Nahuatl#Writing); the code point is just a code point. That last row also uses {{transliteration}} to wrap dotted circle with U+0300: COMBINING GRAVE ACCENT and with U+0302: COMBINING CIRCUMFLEX ACCENT. There is no need to use {{transliteration}} in those places. Consider using {{char}} instead:

{{char|◌}} → ◌ – dotted circle

{{char|ʼ}} → ʼ – modifier letter apostrophe

{{char|◌̀}} → ◌̀ – dotted circle + combining grave accent

{{char|◌̂}} → ◌̂ – dotted circle + combining circumflex accent

—Trappist the monk (talk) 15:08, 28 December 2024 (UTC)

@Trappist the monk It's not a good idea to use Unicode character names as a guide for usage, because they're often inaccurate (or outright wrong), and Unicode will never change them on principle, to ensure stability. In transcriptions of Arabic and Hebrew, ʿ and ʾ perform the role of letters, despite their appearance (e.g. إِسْلَام is transliterated ʾislām, where ʾ represents the consonant [ʔ] and is not modifying anything. Theknightwho (talk) 13:09, 1 January 2025 (UTC)

Unicode is just an implementation detail and your "there is nothing to modify so it is meaningless" argument is obvious BS. Can you please get off your Roman high horse and just fix all these errors you caused making articles across the encyclopedia unreadable then, like at ISO 259? Hftf (talk) 00:00, 30 December 2024 (UTC)

I have to agree that Error: {{Transliteration}}: transliteration text not Latin script (pos 1) is not a helpful error for the transliteration template to raise. The check and error message should be removed.

It is not a requirement for every symbol in every possible transliteration string to be a Latin character. This is demonstrated by the Semitic ʿ (Ayin) which is the recommended transliteration in Unicode.

The Ayin article states "not used as a modifier letter but as a full grapheme", and the Unicode spec dealing with these characters Spacing Modifier Letters. The Unicode Consortium, 1991–2022 has various notes about how numerous named "modifier" letters are used as standalone graphemes, so the 'modifier' argument above is nonsense. Those characters are being used in a transliteration context in a template supposedly designed to enable transliteration markup. There are probably other examples of spacing characters and other punctuation used in transliterations where it makes little sense to discuss whether they are 'Latin' or not. Even if we assume 'romanization' is an exact synonym for 'transliteration', it doesn't necessarily follow that every single glyph used must be 'Latin' by some arbitrary definition.

A transliteration template that cannot accept {{transliteration|ar|ʾ}} or {{transliteration|ar|ʿ}} is broken.

Similarly the examples given above of now broken Scythian languages and Etruscan transliteration templates show that Greek characters are frequently used as 'you know ... "normal"' symbols in various transliteration schemes for ancient languages. I'm not sure this current template supports transliterating various ancient Semitic languages to say Arabic or Hebrew, which might be useful in some contexts, but that's a different issue with the assumed and unspecifiable one-size-fits-all target. Ι think it is reasonable to call semφalχ a 'romanization' of the original Etruscan script, even though it contains unambiguously Greek characters, which are there deliberately. It is definitely a transliteration, using a scheme that pre-dates Wikipedia and Unicode, and possibly IPA. I strongly believe the use of IPA symbols for Etruscan is incorrect, or at least unnecessary, and is just a work around to stop broken-template noise. I don't know much about Scythian, but the script theta seems important and correct. The fix is not to invent a Latin-script-theta, but to accept real world transliterations that exist on and off Wikipedia in the template. This is different from the "Cyrillic small o-with-combining-macron" issue for Japanese which I agree is wrong. The template statement "This template is used to mark up text transliterated or romanised from a non-Latin alphabet script to Latin alphabet script." over-simplifies the definition of what transliteration is, as evidenced by the real-world transliterations that don't conform to this exactly. Most sensible definitions allow for the inclusion of other symbols to enable all relevant parts, given the source context. A heuristic to catch misuse of the transliteration template needs to be smarter, and it seems like there are many ways this this template has been misused caused by the unfortunate 'trans' /literate /late abbreviation confusion. That's also a different problem. The template should not error on valid transliteration schemes. Maintaining a list of lets-call-them-Latin-to-avoid-wiki-errors characters for all possible schemes lumped together is also a waste of time and won't add value. Can we please clear up any remaining confusion about the role and definition of {{transliteration}} and get a less parochial template back? Salpynx (talk) 11:26, 30 December 2024 (UTC)

Paragraph breaks are good. Please use them.

Category:Transliteration template errors has about 130 articles where ʿ (U+02BF MODIFIER LETTER LEFT HALF RING) is the only character wrapped in {{transliteration}}. Most of those contain constructs like this (from Dioceses of the Church of the East after 1552):

[[Shemon VII Ishoyahb|Shem{{transl|ar|DIN|ʿ}}on VII Isho{{transl|ar|DIN|ʿ}}yahb]]

Surely you don't mean to say that this is correct use of {{transliteration}}. Shouldn't that have been written:

[[Shemon VII Ishoyahb|{{transl|ar|DIN|Shemʿon VII Ishoʿyahb|italic=no}}]] → Shemʿon VII Ishoʿyahb

There are Latn-script versions of many Grek-script characters. Your example:

{{transliteration|ett|semφalχ}} → semφalχ

should be rewritten with the Latn-script characters:

{{transliteration|ett|semɸalꭓ}} → semɸalꭓ

—Trappist the monk (talk) 17:55, 31 December 2024 (UTC)

No, they should not "be rewritten with Latn-script characters". Transcribing Etruscan words using Greek phi codepoints was never a problem in the real world until you unilaterally decided to make it one in 2024. And now it inspires ruinous template-removing edits with misleading summaries like these "must". Notice zero google hits when you search the internet for your invented version, compared to the established version? The change brings infinitely negative value.

Instead of searching for the few individual instances of Cyrillic o with combining macron or {{transl|only a single character of a word}} to fix, you're playing dumb and deciding to break thousands of other pages just to make a WP:POINT. This is a serious problem. If you continue this dense denialism, deflection, and passive-aggression, I predict that the next stop is going to be ANI again. Hftf (talk) 03:33, 1 January 2025 (UTC)

I agree with User:Hftf on the Etruscan. Sigmas occur in some Etruscan transliterations too; your suggestion doesn't even work for all cases. Getting into the details of specific transliterations is adding noise to the pretty simple fact that the restriction you added is arbitrary, unhelpful, and wrong. You should justify your change, or revert it.

Yes, those Arabic examples are wrong because the majority of the transliteration is outside the template. The current error message has no bearing on the actual problem, nor the correct fix. I was trying to provide valid transliteration examples that should be accepted by the template but aren't with the made-up "must be Latin" check for every symbol. Salpynx (talk) 04:44, 1 January 2025 (UTC)

Thank you for confirming that the singleton {{transl|ar|DIN|ʿ}} usage I demonstrated above is wrong. Contrary to the statement: The current error message has no bearing on the actual problem, nor the correct fix, the error message bears directly on that problem. Without the error message, no one would know that the error exists. I'll start fixing them.

Let us not exaggerate. ...deciding to break thousands of other pages.... The module change that enabled {{transliteration}} to emit transliteration text not Latin script messages was this 2024-11-28 edit. Following that edit, Category:Transliteration template errors never listed more than a thousand articles. This archive.org snapshot from 2024-12-02 lists 879 articles with the number declining in subsequent snapshots: 2024-12-13, 2024-12-14, and 2024-12-25. As I write this, Category:Transliteration template errors lists ~650 articles.

I've started collecting lists of transliterated singletons and associated language tags that cause Module:Lang to emit the transliteration error. I am also collecting lists of non-Latn-script characters and language tags that appear in (apparently valid) romanizations. With these data I can then modify Module lang to accept these characters. If you know of non-Latn-script characters that are valid transliterations for particular language tags, please list them here; Grek-script φ and χ for ett are already on my list.

—Trappist the monk (talk) 17:14, 1 January 2025 (UTC)

Thank you for confirming that you've left thousands of error messages on hundreds of articles for over a month and are still not doing anything about it. I find it totally unacceptable to leave a visible error message for something that is not even a real error unfixed after 48 hours. Scythian language still has 39 errors, Semitic languages has 30. Using some basic haggadic math then, with one infraction per error message per article per every 48 hours, you've made this encyclopedia suffer over a million times.

This is enough. You're still not getting it. Nobody wants this. Your methods of "break articles first, ask questions later" and "people should come to me begging to add other acceptable characters to the whitelist every time" are both unsustainable and disruptive to make a point. Revert to the status quo before 11/28 or whatever now. Hftf (talk) 19:42, 1 January 2025 (UTC)

I have whitelisted these characters so that Module:Lang accepts them in transliterations:

• ʻ U+02BB: MODIFIER LETTER TURNED COMMA
• ʼ U+02BC: MODIFIER LETTER APOSTROPHE
• ʾ U+02BE: MODIFIER LETTER RIGHT HALF RING
• ʿ U+02BF: MODIFIER LETTER LEFT HALF RING
• Δ U+0394: GREEK CAPITAL LETTER DELTA
• α U+03B1: GREEK SMALL LETTER ALPHA
• β U+03B2: GREEK SMALL LETTER BETA
• γ U+03B3: GREEK SMALL LETTER GAMMA
• δ U+03B4: GREEK SMALL LETTER DELTA
• θ U+03B8: GREEK SMALL LETTER THETA
• φ U+03C6: GREEK SMALL LETTER PHI
• χ U+03C7: GREEK SMALL LETTER CHI
• ϑ U+03D1: GREEK THETA SYMBOL
• ь U+044C: CYRILLIC SMALL LETTER SOFT SIGN
• ᾱ U+1FB1: GREEK SMALL LETTER ALPHA WITH MACRON
• ῑ U+1FD1: GREEK SMALL LETTER IOTA WITH MACRON
• ῾ U+1FFE: GREEK DASIA
• 上 U+4E0A: [CJK Unified Ideographs]
• 入 U+5165: [CJK Unified Ideographs]
• 去 U+53BB: [CJK Unified Ideographs]
• 平 U+5E73: [CJK Unified Ideographs]

See Scythian languages and Semitic languages. The latter still has errors because those {{transliteration}} templates were malformed. Before this edit, the templates had this form – a mix of Ethiopic script and Latin script.

{{transliteration|sem|አነ ʔana}} → [አነ ʔana] Error: {{Transliteration}}: transliteration text not Latin script (pos 1) (help)

The editor's attempted fix didn't fix:

{{transliteration|sem|አነ|ʔana}} → [ʔana] Error: {{Transliteration}}: unrecognized transliteration standard: አነ (help)

Perhaps what is desired is this or sommat similar:

{{lang|sem|አነ}} {{transliteration|sem|ʔana}} → አነ ʔana

Another editor made this 'fix' which should be checked. The need for that 'fix' arose from this January 2020 edit.

—Trappist the monk (talk) 23:36, 1 January 2025 (UTC)