Page 1 of 1

Bern-cursive - my joined-up font

Posted: Fri Aug 21, 2020 6:55 pm
by honest.bern
My aim was to imitate my joined-up handwriting as a boy. (My current handwriting is rather different.) The ascenders and descenders are childish and match the x-height. The font is monoline to imitate a ballpoint pen. I do not intend inflicting this font on the wider world.

I wanted to explore contextual alternates. The font is ugly partly because, lacking the skill to draw, I used geometric shapes, and partly because the original handwriting was ugly. My next font will involve me learning to draw glyphs properly.

Features of the font:
When I link into /e I use a smooth curve. In other joined-up fonts the line changes direction as it passes from link to letter.
The tail of /f curves clockwise, not anticlockwise.
The tail of /q is round.
My primary-school teachers taught a /b open at the top and a /p open at the bottom
The link from below into /m /n /p /r /u /y has a curved top. This is what I learned at school. This feature doubles the number of glyphs for these letters. (See “u.pdf”. I have used different colours for the letters, to show this.)
Again, the round tops of connected /b /h /k /l doubled the number of glyphs.

The lessons that I have learned:

I realised from the start that it would be important to analyse the letter shapes before starting work and to name all the glyphs systematically. Writing a thousand lines of OpenType code sounds daunting, but I have found it easy to edit for three reasons:
(1) the code is actually easy to write,
(2) I organised the code properly, and
(3) my glyph names are as short as possible.
In reality, as I was working, I discovered that my analysis of the original model was faulty and I had to alter the number of letter classes several times.

When I showed the font to my family I discovered that non aficionados have very different views of fonts. I found myself saying “The q (or f) looks like that because this font is based on my handwriting, not yours. If I made a font based on your handwriting then everyone else could complain that some letters were ‘wrong’.”

Before drawing any glyphs I had to decide what to do with the link between letters. My answer was to attach the link to the first of the two letters. Adding the link to the second letter would also have been workable. (Adding half of the link to each letter would have been horrible, and would have increased the number of glyphs needed.) I expect that it would have been possible to use some glyphs for letters (without links) and other glyphs for links (without letters), but I did not try this.

The most extreme example of the link is that going in to /e. (See “e.pdf”.)

The letters /j and /f have tails to the left of the left side-bearing. I added word-initial glyphs so that the tail is not cut off at the start of a line.

Reproducing a historic script involves deciding whether, say, to add a € symbol to a script used before 1995. In this case the problem is /q. The letter as I learned it was designed to be used only before /u. Nowadays I want to be able to join it to other letters. The result is a bodge job.

It is very tempting to make the font look nicer: more like a normal font with inktraps etc. I had to resist this.