Designing Latin S

With Latin S I proposed a new standard Latin character set that supports more than the European languages. Some of the additional characters might be unknown to type designers though, so here’s a closer look.

Overview

Additional characters to the common practice are marked green (currency symbols yellow, and the red ones might be omitted):

That makes about 80 additional glyphs – which serve more than 200 million addiotional speakers!

The example images below show the glyphs as sans serif (Pangea Text) and serif (Brill) designs.

◌̣

Char	Unicode	AGLFN name
̣	0323	dotbelowcomb

Ok, let’s start with an accent. You probably already have ◌̇ (/dotaccentcomb), so you just need to duplicate it and move the outline below the baseline.

ẠḄḌẸḤỊṆỌṢỤẒ ạḅḍẹḥịṇọṣụẓ

Show character table

Char	Unicode	AGLFN name
Ạ	1EA0	uni1EA0
Ḅ	1E04	uni1E04
Ḍ	1E0C	uni1E0C
Ẹ	1EB8	uni1EB8
Ḥ	1E24	uni1E24
Ị	1ECA	uni1ECA
Ṇ	1E46	uni1E46
Ọ	1ECC	uni1ECC
Ṣ	1E62	uni1E62
Ụ	1EE4	uni1EE4
Ẓ	1E92	uni1E92
ạ	1EA1	uni1EA1
ḅ	1E05	uni1E05
ḍ	1E0D	uni1E0D
ẹ	1EB9	uni1EB9
ḥ	1E25	uni1E25
ị	1ECB	uni1ECB
ṇ	1E47	uni1E47
ọ	1ECD	uni1ECD
ṣ	1E63	uni1E63
ụ	1EE5	uni1EE5
ẓ	1E93	uni1E93

Now you can easily create these ones!

ǍẼḠǦḦǏṄǑṖǓẌỲỸȲ/Jacute ǎẽḡǧḧǐṅǒṗǔẍỳỹȳ/jacute

Show character table

Char	Unicode	AGLFN name
Ǎ	01CD	uni01CD
Ẽ	1EBC	uni1EBC
Ḡ	1E20	uni1E20
Ǧ	01E6	Gcaron
Ḧ	1E26	uni1E26
Ǐ	01CF	uni01CF
Ṅ	1E44	uni1E44
Ǒ	01D1	uni01D1
Ṗ	1E56	uni1E56
Ǔ	01D3	uni01D3
Ẍ	1E8C	uni1E8C
Ỳ	1EF2	Ygrave
Ỹ	1EF8	uni1EF8
Ȳ	0232	uni0232
	None	J.latnNLD
ǎ	01CE	uni01CE
ẽ	1EBD	uni1EBD
ḡ	1E21	uni1E21
ǧ	01E7	gcaron
ḧ	1E27	uni1E27
ǐ	01D0	uni01D0
ṅ	1E45	uni1E45
ǒ	01D2	uni01D2
ṗ	1E57	uni1E57
ǔ	01D4	uni01D4
ẍ	1E8D	uni1E8D
ỳ	1EF3	ygrave
ỹ	1EF9	uni1EF9
ȳ	0233	uni0233
	None	j.latnNLD

You already have all the base letters, you already have the accents, now compose these!

ʼʻ

Char	Unicode	AGLFN name
ʼ	02BC	uni02BC
ʻ	02BB	uni02BB

Just duplicate /quoteleft/quoteright – done!

Congrats, you already completed more than 60% of the additional glyphs! 🎉

Ɲɲ

Char	Unicode	AGLFN name
Ɲ	019D	uni019D
ɲ	0272	uni0272

If your /N and /J are ready, just combine those two.

Ʌʌ

Char	Unicode	AGLFN name
Ʌ	0245	uni0245
ʌ	028C	uni028C

You aready have these glyphs, only turned. 🙃

Ɔɔ

Char	Unicode	AGLFN name
Ɔ	0186	uni0186
ɔ	0254	uni0254

Turning /C and /c might do the trick here, too!

Ɛɛ

Char	Unicode	AGLFN name
Ɛ	0190	uni0190
ɛ	025B	uni025B

Numeral 3 might be a good starting point for these, or you re-use the Cyrillic letters Зз (/Ze /ze) or the Greek ε (/epsilon)

Ǝǝ

Char	Unicode	AGLFN name
Ǝ	018E	uni018E
ǝ	01DD	uni01DD

Mirroring /E and turning /e will probably all you need to do about these two!

Ɨɨ

Char	Unicode	AGLFN name
Ɨ	0197	uni0197
ɨ	0268	uni0268

Just add strokes to /I and /i. Maybe have a look at your /Eth and /dcroat.

ƁƊƘƳ ɓɗƙƴ

Char	Unicode	AGLFN name
Ɓ	0181	uni0181
Ɗ	018A	uni018A
Ƙ	0198	uni0198
Ɲ	019D	uni019D
Ƴ	01B3	uni01B3
ɓ	0253	uni0253
ɗ	0257	uni0257
ƙ	0199	uni0199
ɲ	0272	uni0272
ƴ	01B4	uni01B4

Now for some curls, and you’re nearly done!

ẞƏə

Char	Unicode	AGLFN name
ẞ	1E9E	uni1E9E
Ə	018F	uni018F
ə	0259	uni0259

Uppercase Eszett and the Latin Schwas are already included in many fonts but I added them for the sake of completeness.

ə is a mirrored /e, the uppercase variant is basically the same, only scaled. If you already designed the Cyrillic Ә & ә: They’re identical.

On Typefacts I wrote some more about the German Capital Letter Eszett.

Ŋ

Char	Unicode	AGLFN name
Ŋ	014A	Eng

There are two shape variants for the uppercase /Eng: One preferred in Sami languages (based on the uppercase /N) and one preferred in Afrikan languages (looking like the lowercase /n). (Brill even has three different shapes, see image)

I’d recommend the latter one as default since it is used by many more speakers than the Sami shape.

You might want to draw both shapes thought and substitute them in the locl feature, see Latin Character Sets > Eng.

Ŀŀ

Char	Unicode	AGLFN name
Ŀ	013F	Ldot
ŀ	0140	ldot

These only occur in Catalan in the combinations /L/periodcentered/L and /l/periodcentered/l. I recommend using unencoded glyphs instead, in combination with the locl feature, see Latin Character Sets > L_periodcentered.latnCAT.

₵₡₲₺₼₦

Char	Unicode	AGLFN name
₵	20B5	uni20B5
₡	20A1	colonmonetary
₲	20B2	uni20B2
₺	20BA	uni20BA
₼	20BC	uni20BC
₦	20A6	uni20A6

If your font supports certain languages, you might consider also supporting the currency symbols of the countries where they are spoken:

• ₵: Cedi (Ghana)
• ₡: Colón (Costa Rica)
• ₲: Guaraní (Paraguay)
• ₺: Lira (Turkey)
• ₼: Manat (Azerbaijan)
• ₦: Naira (Nigeria)

IJijŦŧĸʼnſ

Char	Unicode	AGLFN name
IJ	0132	IJ
ij	0133	ij
Ŧ	0166	Tbar
ŧ	0167	tbar
ĸ	0138	kgreenlandic
ʼn	0149	napostrophe
ſ	017F	longs

You won’t necessarily need these glyphs from Unicode range Latin Extended A.

• IJ, ij: If you want to draw these ligatures, I recommend using unencoded glyphs in combination with OpenType features
• Ŧ, ŧ are needed for Sami languages only (≈25,000 speakers) or Balante language in Senegal (≈100,000 speakers)
• ĸ is a historical letter used in Greenlandic until the 1970s and Labrador Inuttut until the 1980s
• ʼn is deprecated and shouldn't be used
• ſ is an historical form of s

Kerning

In most of the cases you can reuse kerning classes for your existing character set, eg.:

Char	Left Kern Group	Right Kern Group
Ʌ	A	A

Kern ƁƊ

Char	Left Kern Group	Right Kern Group
Ɓ	Hook	B
Ɗ	Hook	O

The minimum Kerning:

Kerning Pair	Example Word	Translation
AƁ	GÁAƁÀA	the shore
KƁ	-	-
LƁ	-	-