Private Use area vs OpenType Features?

[Julian Griffin]

Hello,

  I am new to OpenType Features and I am confused as to when it is best to use them. I am hoping if I explain what I am doing, someone can tell me if I should go ahead with what I have already done or should I persevere with learning how to use OpenType Features.

I have created:
  the 26 lowercase Latin letters (u+0061-u+ 007A),
  the Combining Macron (u+0304),
  the Combining Dot Above (u+0307)
  the Combining Double Macron (u+035E),
  the Combining Double Inverted Breve (u+0361),
  the Combining Long Solidus Overlay (u+0338),
  the Combining ‘x’ above (u+033D)

Then in the Private Use area of Unicode (u+F000-u+FFFF) I have created 46 graphemes consisting of:
  25 lowercase letters (u+0061-u+ 007A),
  5 single lowercase letters each with a Combining Macron (0304),
  4 sets of two lowercase letters each set with a Combining Double Macron (u+035E),
  12 sets of two lowercase letters each set with a Combining Double Inverted Breve (u+0361)

I have then used these 46 graphemes as the basis for 46 superscript graphemes, which are shifted left 1/2 en. All created in the Private Use area.

I have then used these 46 superscripted graphemes as the basis for four sets of 46 graphemes each set shifted a little further left. Also all created in the Private Use area.

The Combining Dot Above (u+0307) is used to overlay the 5 lowercase vowels in the Private Use area. These 5 dotted vowels have not been created in the Private Use area

The Combining Long Solidus Overlay (u+0338) and the Combining ‘x’ above (u+033D) are used to overlay the 52 Latin letters. These have also not been created in the Private Use area.

So I have created all these 276 graphemes listed above in the Private Use area.

So finally to my question; should I continue the use the Private Use area or should I persist in learning about OpenType Features and what are the advantages of each? And a sub-question; if OpenType Features is the way to go, which feature(s) should I use for what I have outlined above?

Thank you for your help, best regards,
Julian

[Bhikkhu Pesala]

5 single lowercase letters each with a Combining Macron (0304),

These characters exist in the Latin Extended characters sets: āēīōū (0101, 0113, 012B, 014D, 016B). They don't need to be in the Private Use Area.

The Combining Dot Above (u+0307) is used to overlay the 5 lowercase vowels in the Private Use area. These 5 dotted vowels have not been created in the Private Use area

These characters also exist: ȧ (0277) ė (0117) ȯ (022F). If characters have an assigned codepoint it is always best to use it rather than the PUA, but not all characters will have an assigned codepoint. What would lowercase i with dot above look like if not i ?

So finally to my question; should I continue the use the Private Use area or should I persist in learning about OpenType Features and what are the advantages of each? And a sub-question; if OpenType Features is the way to go, which feature(s) should I use for what I have outlined above?

Glyphs that are not available in the Unicode characters sets can be mapped to the Private Use Area, or they can left unmapped if they are to be used by OpenType features.

The first decision that needs to be made is how are users going to type these double macron, solidus, and dot above characters? Wouldn't they be better off typing the two ordinary vowels then the combining double macron? au͞ ao͞

I presume this font has something to do with phonetics, a subject that I am unfamiliar with, so I have no idea how people usually typeset this kind of work. A PDF file of some existing work might give us an idea how to proceed.

Selecting glyphs from a character map is shockingly inefficient. I use a customised Windows keyboard to type ā, ē, ī, ō, ū and Ā Ē Ī Ō Ū (and other characters required for Pāḷi or Sanskrit transliteration) very easily. It would not be too hard to do something similar for combining double macron, or combining solidus or dot above.

If the font requires OpenType feature support, that already limits its usefulness to a few supporting applications. I don't have Word, but I believe it's OpenType support is quite limited.

[Julian Griffin]

Thank you Bhikkhu Pesala for your response.

  You are right I am working on a phonic alphabet. Unlike other phonic systems this one can be used to mark-up Standard English text with phonic clues as to how to read a word. It also must have structure and order so word-lists can be sorted. Along with this I am working on a phonic dictionary so words can be looked up as they are spoken not how they are spelt. So it is imperative for a recognisable and predictable sort order. That is the reason for using the Private Use area.

  I am going to use Microsoft’s MSKLC to aid with grapheme entry.

  But I think you have given me the answer I needed, for a font to be universally usable I should continue with a CodePoint per-glyph I have already created.

So again, thank you for your help.
Julian :~)