Louie Mantia, Jr.Hey pals! I’m gonna do a short little video series here on the OpenType features inside Womprat so you can learn how to use them!
Buy the font here → http://womprat.xyz/
First up— Ligatures.
Next— Contextual Alternates.
Onto Discretionary Ligatures!
Arbitrary fractions. Neat!
Alright— Ordinals. Those little letters that appear after numbers.
This is where the fun begins!
SS01: Special Edition
SS02: Empire
SS03: Jedi
SS04: Squadron
Wait— Stylistic Sets are not mutually exclusive?!
SS05: Desiljic
SS06: Wanga
SS07: X-Wing
SS08: Mortis
SS09: Director’s Cut
SS10: Youngling
...we’re now halfway through the Stylistic Sets.
Yeah man, this font is crazy.
Gonna take a little break before I do the second half. Thanks for watching and learning about Womprat (the most complicated font I’ll ever make)!
Also if you wonder *how* all of this works, @AurekFonts wrote extensive code for all these features into the font.
Okay! Back at it.
SS11: Lubalin Ligatures
SS12: Avant
SS13: Vavnt
SS13: Vavnt
Reminder! You can stack Stylistic Sets.
SS14: Roman Numerals
Ender (@AurekFonts) engineered all of the stuff you see in these videos, but this one is especially fun.
SS15: Catchwords
An addendum to Catchwords—
SS16: Bold Punctuation
SS17: Heavy Quotes
SS18: Border Brackets
Chuck Skoda@louie @AurekFonts this is mental to me. I had no idea fonts could do this. Is there like a table in the font of all reasonable Roman numeral combinations? Is there executable logic happening? As you’ve gone through this, my mental conception of fonts has been BLASTED INTO OBLIVION (<- would probably look great in Womprat).
AurekFontsThis is a good question! The basic principle is to use "contextual lookups"—basically, the font itself checks the context around a symbol and decides whether to replace it.
In this case, I created three groups (lists) of glyphs:
- Letters to replace with roman numerals
- Roman numerals
- & EVERY OTHER LATIN/GREEK/CYRILLIC LETTER
1/
Now the next part went through some drafts. The first looked something like this:
- Find a letter that could be a roman numeral
- Check the surrounding context (forward and backwards), to see if the current word contains any letters that cannot become roman numerals; ignore those
- If none of the "ignore" conditions are met, replace with a roman numeral
2/
The next draft (the one we actually used) might be a little less efficient, but it's less verbose.
(Also, due to a weird technical constraint, this version worked better on our font tester webpage.)
This one is similar to the last but does things in a different order, and swaps the letters twice:
- Replace every letter with the roman numeral equivalent
- Then check context to replace the roman numerals with letters again
3/
There are some limitations to this implementation, so this is something we may update in a future version:
- There is no check for the ORDER of the roman numerals, so you can type invalid sequences that still get converted
- There are a handful of words in English that only contain roman numeral candidates, which will thus be replaced with roman numerals, even though they are not.
4/
@AurekFonts @louie Definitely going to need to read up on this. I'm confused at what this line does:
ignore sub @noRoman @noRoman @noRoman @noRoman @noRoman @noRoman @noRoman @noRoman @noRoman @toRoman'
That this one doesn’t:
ignore sub @noRoman @toRoman'
Doesn't one instance of @toRoman’ preceded by @noRoman ignore the same sub? At first glance, I would think you'd only want the first entry on the top half of these lines to be @noRoman, and the rest to be @toRoman until the final @toRoman’
Markus Trautwein@louie I love your work on that.
Luke Dorny 
@louie most impressive. Very satisfying to see your work in action and see the how and why. 👍👍