Sources

Standards

• Unicode Technical Report #25: Unicode Support for Mathematics

  The Unicode report on mathematics (revision 15). It explains what the Mathematical Alphanumeric Symbols are for and says they should not replace markup for styling text.

  Used for: Mathematical Alphanumeric Symbols are meant for mathematical notation and should not be used as a substitute for styling markup.

• Unicode FAQ: Ligatures, Digraphs, Presentation Forms vs. Plain Text

  The Unicode FAQ on ligatures, digraphs and presentation forms. It says the mathematical alphanumeric symbols are meant for mathematics and calls their use for styled-looking plain text an unintended side effect.

  Used for: Using the mathematical alphabets to imitate styled text in plain text is an unintended side effect; ordinary emphasis belongs in markup.

• Unicode Character Database: UnicodeData.txt

  UnicodeData.txt lists reserved holes in U+1D400 to U+1D7FF and the Letterlike Symbols characters that fill them.

  Used for: Some styled letters live outside the Mathematical Alphanumeric Symbols block because they were encoded earlier (for example script B, Fraktur C, double-struck R, italic h).

• Unicode names list: Mathematical Alphanumeric Symbols

  Names list for U+1D400 to U+1D7FF including notes pointing to Letterlike Symbols for the holes.

  Used for: The Mathematical Alphanumeric Symbols chart documents each styled alphabet and its cross references.

• Unicode Standard Annex #15: Unicode Normalization Forms

  The Unicode standard for normalization forms (revision 58), including the compatibility forms NFKC and NFKD. The inspector offers NFKC as an optional extra step after its exact mapping.

  Used for: Compatibility normalization (NFKC and NFKD) removes many formatting distinctions, can remove distinctions that matter to the meaning of text, and must not be applied blindly to arbitrary text.

• Unicode Technical Standard #39: Unicode Security Mechanisms

  Unicode Technical Standard #39, Unicode Security Mechanisms, defines confusable characters and mixed-script detection.

  Used for: Characters from different scripts can look identical (confusables), which matters for security and impersonation.

Platform help pages

• Facebook Help Center: Names allowed on Facebook

  Facebook's help page on the names it allows, with a list of what a profile name cannot include.

  Used for: Facebook profile names cannot include symbols, numbers, unusual capitalization, repeating characters or punctuation, or characters from multiple languages.

• Discord Support: New Usernames & Display Names

  Discord's article on usernames and display names, with the length and character rules for each. Display names may also use special and non-Latin characters, within the Community Guidelines.

  Used for: Discord usernames are 2 to 32 characters, lowercase, and limited to Latin letters, numbers, underscore, and period.

  Used for: Discord display names are 1 to 32 characters and may include special characters, spaces, emoji, and non-Latin characters.

• Discord Support: Markdown Text 101

  Discord's Markdown guide lists the chat formatting syntax for bold, italics, underline and strikethrough.

  Used for: Discord messages support Markdown formatting: **bold**, *italics*, __underline__, and ~~strikethrough~~.

• Instagram Help Center: Add a bio to your Instagram profile

  Instagram's help page on adding a bio to a profile, including the length limit.

  Used for: An Instagram bio can be up to 150 characters.

• TikTok Help Center: Changing your username

  TikTok's help page on changing your username gives the length, the allowed characters and how often it can change.

  Used for: TikTok usernames are 2 to 24 characters, can only contain letters, numbers, underscores and periods, cannot end with a period, and can be changed once every 30 days.

• TikTok Help Center: Changing your nickname

  TikTok's help page on changing your nickname gives the 30 character limit and the 7 day wait between changes.

  Used for: A TikTok nickname can be up to 30 characters and changed once every 7 days.

• X Developer Platform: Counting characters

  X's developer documentation on counting characters, with the 280 limit, the weights, emoji, links and NFC.

  Used for: An X post can have up to 280 characters. Text is normalized to NFC; Latin letters and common punctuation count as 1, emoji, CJK characters and other Unicode count as 2, and every link counts as 23.

Technical documentation

• MDN: Intl.Segmenter

  Documents the grapheme granularity of Intl.Segmenter.

  Used for: Intl.Segmenter with grapheme granularity splits text into user-perceived characters.

• twitter-text: counting configuration (config/v3.json)

  The configuration file of twitter-text, X's open-source counting library, with the code point ranges that count as 1, the default weight of 2 and the 23-character link length.

  Used for: Code points U+0000 to U+10FF, U+2000 to U+200D, U+2010 to U+201F and U+2032 to U+2037 count as 1 on X; all others count as 2.

Community evidence

• WebAIM discussion list: Unicode styled letters and screen readers

  Accessibility practitioners on the WebAIM list report that styled math letters are read by name or poorly. Behavior varies by screen reader and version.

  Used for: Screen readers may announce styled Unicode letters by their character names, spell them out, or skip them, depending on the reader and its settings.