TUS Section 2.11 says, "If the combining characters can interact typographically—for example, U+0304 combining macron and U+0308 combining diaeresis — then the order of graphic display is determined by the order of coded characters (see Table 2-5). By default, the diacritics or other combining characters are positioned from the base character’s glyph outward".
So, if I have two spacing combining marks E and O that are each positioned to the left of the base (say X) in a left-to-right script, so that the encodings <X, E> and <X, O> appear with the glyph orders <gE, gX> and <gO, gX>, and codings <X, E, O> and <X ,O, E>, if not total gibberish, represent a horizontal sequence of the glyphs with gX on the right, should <X, E, O> render as <gE, gO, gX> or <gO, gE, gX>? The phonetics and collation (in so far as it is meaningful) of the words provide no cue as to the order of the encoded characters. I have encountered both renderings. The issue came up when I was checking, in both the Firefox and MS Edge browsers, that my OpenType Tai Tham font Da Lekh could handle all the headwords of two Northern Thai dictionaries. (Sparing dotted circle deletion and orthographic syllable reunification are tricky.) One of the dictionaries spells a few words with a combination of the Tai and Pali notations for the vowel /o:/ in open syllables where one might expect to see an independent vowel. I'm down to two other rendering engine issues - a combination of tone mark and then vowel in 4 words, where the dictionary probably has a misspelling, and the need for an OpenType feature (probably a cvXX) for inconsistent handling of U+1A58 MAI KANG LAI. The latter may be a challenge - I couldn't persuade MS Edge to use the font's Lao shaping for the Tai Tham script or for the Latin script in a transliteration mode. (That mode is triggered by feature ss02 for the Latin script, and that works well enough in browsers.) Richard.
