Talk:Unicode subscripts and superscripts

When to use

I agree the second para needs a citation; I can't find one off hand. There are clearly pros and cons - interop is easier without markup, but the range of values is smaller. Will continue to look for some citations, maybe there was once some discussion in the WG, though I dont know these are public. Justinc (talk) 11:48, 13 September 2010 (UTC)

I don't know any current recommendations that make special exemptions for "²" and "³". That may be a leftover from the time when common fonts and computing platforms in English-speaking countries had trouble coping with characters outside the Latin 1 repertoire. But those days are mostly gone. So I replaced the paragraph with some quotations from Unicode in XML and other Markup Languages. Indefatigable (talk) 22:30, 14 September 2010 (UTC)

	markup			unicode
copied	H2O	32	51⁄4	H₂O	3²	5¹⁄₄
pasted	H2O	32	51⁄4	H₂O	3²	5¹⁄₄

For anyone considering whether to use these or not: The Unicode Consortium and current HTML/CSS guidelines recommend using markup (such as <sup>) which will look more consistent. However if the reader copy/pastes text with markup they will often lose said markup when they paste the text. ―MJBurrage^(T•C) 16:21, 8 January 2016 (UTC)

Table syntax

Is there a reason why the table is in HTML, not in wiki syntax? --92.226.196.72 (talk) 23:21, 23 February 2008 (UTC)

0 1 2 3 4 5 6 7 8 9 A B C D E F U+207x x⁰ xⁱ     x⁴ x⁵ x⁶ x⁷ x⁸ x⁹ x⁺ x⁻ x⁼ x⁽ x⁾ xⁿ U+208x x₀ x₁ x₂ x₃ x₄ x₅ x₆ x₇ x₈ x₉ x₊ x₋ x₌ x₍ x₎   U+209x xₐ xₑ xₒ xₓ xₔ

0 1 2 3 4 5 6 7 8 9 A B C D E F U+207x x0 xi     x4 x5 x6 x7 x8 x9 x+ x− x= x( x) xn U+208x x0 x1 x2 x3 x4 x5 x6 x7 x8 x9 x+ x− x= x( x)   U+209x xa xe xo xx xə

No, and I replaced it with your version. babbage (talk) 20:16, 3 October 2010 (UTC)

WIkipedia: ?

Can somebody explain the addition of "Wikipedia:" to the page name?

This does not appear to have been done with other pages describing Unicode subsets.

The move was mistaken, since the topic of which character encoding to use to represent superscripts and subscripts is not Wikipedia-specific, but rather a topic that is relevant to all of digital typography. Thus it is incorrect to say that it's not encyclopedic (which was the argument given for the move). It should be moved back. If one claims that "Unicode subscripts and superscripts" doesn't "count" as a worthy topic for an encyclopedia article, then you have to move half of Wikipedia out of the article namespace into the project namespace. It doesn't pan out upon analysis. — ¾-10 20:50, 10 December 2011 (UTC)

Update: I asked at the help desk for someone to move it back, and it was restored. — ¾-10 17:01, 11 December 2011 (UTC)

Poor wording

"...your browser for instance produces 11⁄12..."

Not everyone reading this will be using a browser. Not everyone reading an encyclopedia expects to be addressed personally nor is it a very appropriate example. Very poor use of language here. — Preceding unsigned comment added by 86.141.142.230 (talk) 18:58, 9 April 2012 (UTC)

Browser support

I am surprised that Google Chome on my tablet (Samsung Tab 2 - GT-N8000) is missing many of these chars, only showing 1-4 in both, while on my phone they are ok. I also tested in the app "Unicode Chars" with the same results. Maybe some fonts need to be installed, but it isnt obvious how to do that.

I dont see the same problem with Firefox on the tab, and no problems with Chrome on Ubuntu. Are there any other browsers with problems on some platforms? John Vandenberg ^(chat) 13:37, 29 June 2016 (UTC)

It makes sense that older mobile devices may lack Unicode support. Maybe the phone is newer and has more Unicode support. Do you need some specific Unicode superscripts and subscripts for some purposes? 2A01:119F:21D:7900:89A7:4322:8B46:1724 (talk) 07:04, 28 June 2018 (UTC)

1120 more superscripts, subscripts and small capitals

An IP editor keeps adding an external link to a document by Piotr Grochowski requesting to add 1120 more superscripts, subscripts and small capitals to the Unicode Standard. I do not believe that it is appropriate to link to this document from this article because even a well-written character encoding proposal is not notable until and unless it is accepted by the Unicode Technical Committee (UTC); and in this case the document is so poorly written and lacking in substance that there is zero chance that the UTC would ever accept it. BabelStone (talk) 11:39, 28 June 2018 (UTC)

That author and some editor here consistently reverses the meaning of "true superscripts" from how most people understand that term. The <sup> is supposed to produce a "true superscript". The fact that many font rendering engines just shrink the full size character is a fault with the font rendering, not with the choice of encoding.

From the Unicode docs it is apparent that they intended these glyphs to also be "true superscripts" and actually indistinguishable from sup/sub markup in a properly-working font engine. However font designers thought these were much more useful for numerator/denominator fractions, which are far worse when attempted with sup/sub markup and thus more important to find a substitute for. This means they are much smaller and closer to the baseline than most fonts want for super/sub scripts. Unicode intended this to automatically happen for numbers separated by the fraction slash, but as far as I can tell almost no font rendering software does this, and it may be a terrible idea as it requires it to parse out exactly what portion of the surrounding text is the numerator and denominator. I expect fixing this fraction rendering will not happen for decades at least after any possible fixing of super/subscript rendering, so the use of these glyphs for numerator/denominator is probably a very good idea even it is not what Unicode wanted.Spitzak (talk) 18:13, 28 June 2018 (UTC)

I found an error in this proposal today. The proposal actually includes 1118 characters, not 1120, as it double counts superscript ¹/superscript superscript 1 and subscript ¹/subscript superscript 1. 124.170.251.21 (talk) 17:46, 3 March 2019 (UTC)

FYI: The Script Ad Hoc group gave L2/18-206 the thumps down in July, 2018 per L2/18-241 so it's safe to say this document does not belong in this article. DRMcCreedy (talk) 19:48, 3 March 2019 (UTC)

Issue discussion is now at github. 60.106.173.239 (talk) 15:24, 13 March 2019 (UTC)

Some examples please

For those of us who are not technical experts it would be a great help to have some examples on how to use these characters. JFB80 (talk) 20:12, 3 February 2019 (UTC)

I'm not sure I understand. They're Unicode characters; you enter them the same ways you enter other Unicode characters. I'm not sure it's appropriate for us to get into more detail then that on this page.--Prosfilaes (talk) 09:36, 5 February 2019 (UTC)

I don't understand what about his comment you don't understand. He's asking for some examples of the intended use cases for these characters. This isn't an unreasonable question, especially given that you obviously cannot super- / subscript any arbitrary text using these characters. They're only intended for particular uses, and showing what these uses are would therefore be very helpful to the average reader. — Preceding unsigned comment added by 92.67.227.181 (talk) 16:44, 2 July 2022 (UTC)

Doesn't the first sentence of the "Uses" section cover this? DRMcCreedy (talk) 21:22, 2 July 2022 (UTC)

Overscript

@Kwamikagami: If "overscript" characters are to be included in the tables I think they should be defined, with references, somewhere in the article. Otherwise it strikes me as original research. DRMcCreedy (talk) 16:22, 30 November 2019 (UTC)

"ꭟ" listed at Redirects for discussion

A discussion is taking place to address the redirect ꭟ. The discussion will occur at Wikipedia:Redirects for discussion/Log/2021 March 19#ꭟ until a consensus is reached, and readers of this page are welcome to contribute to the discussion. 𝟙𝟤𝟯𝟺𝐪𝑤𝒆𝓇𝟷𝟮𝟥𝟜𝓺𝔴𝕖𝖗𝟰 (𝗍𝗮𝘭𝙠) 08:19, 19 March 2021 (UTC)

very poorly supported

This might have been true in 2018, but is this still true? I'm asking because all the examples appear fine to me. 92.67.227.181 (talk) 16:39, 2 July 2022 (UTC)

2022 changes

There seems to have been some changes made to the unicode subscripts and superscripts. These have borked this page and many others across the internet. Can anyone help explain what has changed and how to fix things? Ehdeejay (talk) 21:02, 27 September 2022 (UTC)

What do you mean by "borked"? — kwami (talk) 23:47, 27 September 2022 (UTC)

Greek table seems inconsistent

The character ᵅ (U+1D45 MODIFIER LETTER SMALL ALPHA) is missing from the Greek superscript and subscript letters table.

Perhaps this is because U+1D45 is classified as Latin. However, this is not consistent, as the following Latin characters are currently present in the Greek table:

• ᵋ (U+1D4B MODIFIER LETTER SMALL OPEN E)
• ᶥ (U+1DA5 MODIFIER LETTER SMALL IOTA)
• ᶹ (U+1DB9 MODIFIER LETTER SMALL V WITH HOOK)

Is there a rationale to discriminate U+1D45 apart from these three characters? Ei283 (talk) 07:36, 8 August 2023 (UTC)

It doesn't look like an alpha, whereas the other letters look Greek. I don't know of sources that contrast between Latin and Greek forms for the others, but a distinction is common for script a vs alpha. — kwami (talk) 10:53, 8 August 2023 (UTC)

Perhaps ᵅ does not visually resemble the Greek letter α, as can be decided with the following comparison, with Greek alpha on the left and Latin alpha on the right:

αɑ

Certainly, then, ᶹ (U+1DB9 MODIFIER LETTER SMALL V WITH HOOK) should not represent the Greek letter υ, as the differences are much more apparent, as can be seen with Greek upsilon on the left and Latin V-with-hook on the right:

υʋ

Personally, all of my fonts display Latin and Greek alpha characters identically, so I cannot distinguish them and thus do not see any reason not to use U+1D45 as a Greek superscript alpha. However, several of my fonts draw Greek upsilon and Latin V-with-hook very differently, to the extent that I cannot identify them as the same letter. Ei283 (talk) 20:56, 8 August 2023 (UTC)

My fonts are the opposite: alpha doesn't look like the same letter, but upsilon does. If it's that distinct for you, then I agree that it shouldn't be in the Greek table. — kwami (talk) 21:03, 8 August 2023 (UTC)

The table is inconsistent because it currently uses a "this character looks like a Greek letter" criteria instead of using Unicode properties like script. I'd recommend removing the three Latin characters from the Greek table. DRMcCreedy (talk) 19:59, 8 August 2023 (UTC)

There is a distinction between script a and alpha e.g. in Dania, which uses both. AFAICT there is no distinction for epsilon and iota. The fact that Unicode defines a symbol as being part of a particular script doesn't mean it's not used for other scripts. If you wanted a superscript Greek epsilon, this is the character you'd use. If someone requested Greek variants of these, it's possible Unicode would encode them, but it's also possible they'd instruct people to use the existing characters. But until then, this is what we have. If you really believe we shouldn't provide these for the convenience of the reader, I suppose we could replace them with asterisks and a note that they exist but are defined as Latin. — kwami (talk) 20:54, 8 August 2023 (UTC)

I did that for alpha and upsilon, given the comment above. — kwami (talk) 21:07, 8 August 2023 (UTC)