Re: a question of hyphenation policy

[Robert Thorsby]

Hi again Branden,

I'll let others respond regarding the morality of dangling hyphens on 
the last line. :-)

On 30/8/24 16:38, G. Branden Robinson wrote:
> > At 2024-08-30T16:29:03+1000, Robert Thorsby wrote:
> > I wouldn't call myself a typographer but I refuse to allow hyphens to
> > break, usually via '.hy 0' or by using '\%'. I then kern the tripe out
> > of any offending line, often using Ted Harding's "poor man's kerning"
> > trick.
>
> Ooh, that sounds like something good to remind the youngsters of.  Have
> a link handy?

I've used Ted's "solution" for so long I have forgotten when he posted 
it. It is spectacularly simple, but finicky, to implement. With 
justified text you take a new line in your input, then
\s[-X]\H[+X]text to be kerned goes here\H[0]\s0

Note, the value of X given to the opening pair of escapes must be equal 
but opposite in sign and must include a scaling unit. I always use "u" 
as the unit. Thus
\s[+167u]\H[-167u]

It works simply by reducing/increasing the pointsize by the stated 
amount and then immediately stretching/shrinking the vertical height of 
the type by an opposite amount. The right bookends simply reverse the 
effect.

You can fiddle with the value of X in basic units until you get it 
"exact" or you can get it within a character or two of end-of-line and 
just add "\p". You may have to add a "\p" to the line before you start 
playing around.

I have found that using values of X that bear some relationship with the 
pointsize you happen to be using is easier to work with. In the example 
167u seems to be a good starting point for 12pt type, rather than (say) 
150u.

I have never been able to tell the difference on the printed page 
between the "unkerned" lines above and below and the intermediate 
"kerned" line(s). Of course, the difference is beyond any monitor's 
capabilities.

Robert Thorsby