Font-tastic work

[MikeW]

It takes a font-making wacko like myself to stretch the boundries of OpenType coding--coding that can fail in many applications.

 

I reported a font issue to Dave. It concerns a font that I was making that makes heavy use of chaining contextual alternates that wasn't working quite correctly in AD. The latest beta corrects that problem.

 

So what's a chaining contextual alternate feature you ask? Simply put, it is a set of rules in an OpenType font that in this case, and depending on particular characters that follow other characters, changes those characters because of their context according to a certain historical usage.

 

Huh? The English & French had particular rules governing what is called the long s (a medial s) character. Historically, that medial s character resembles a lowercase f and differs only in the cross-bar bisecting the stem. The rules for the long s usage are somewhat variable and depend on the country and the period of usage. Which makes for a lot of proof reading a historical text that makes use of such characters.

 

Because I am setting a manuscript that quotes a lot of late 18th to early 19th century French writings and designed the font for this purpose, I thought I would give it a go to roll those rules into the font I designed. 

 

Ok, here's the lowercase f we all know and love:

 

 

And here's a long s hardly anyone uses any more:

 

 

Back to those rules. There was always also what we all know as a lowercase s character as well. Whether a lowercase s is used or whether the long s form is used have to follow the period's rules governing their usage.

 

For instance, the lowercase s is used if a lowercase f precedes it. That was to keep confusion down because as you can see from the above screen shots, they look pretty similar. (And don't ask why they had two s characters...). As well, a lowercase s is used if preceding a hyphen or at the end of a word. Another instance is a lowercase s is used if it is preceded by and followed by a long s. And, of course, there are exceptions.

 

Which makes writing rules into a font a difficult task. But insofar as the historical texts are concerned that make up the manuscript, I did succeed. And the job headed off to print next Monday.

 

But then I decided to see what applications the font would work as designed and which wouldn't work correctly in. Which led to finding the issue in AD, and QXP, and ... well, a few applications. So I notified the respective companies and hey, guess what? Serif was the first to roll out the fix.

 

So this text without the historical feature turned looks like (the text is made of real words for testing sake):

 

 

And when the historical feature is turned on in AD, that text transforms into:

 

 

Ok, remember I mentioned there are exceptions? You can see that I haven't figured out how to implement one rule: The ending long s character when a proper noun is abbreviated (the Genes. text above). But that was easy enough to use GREP searches for and change that single character in ID.

 

I am very happy with Serif's responsiveness to this chaining contextual alternates issue. And Dave in particular--thanks, Dave.

 

Mike

[TheaMusselwhite]

Hi Mike

You really seem to know your fonts (and coding for opentype)! I know absolutely nothing about Opentype, and was wondering if I could ask you a quick question?

I haven't yet purchased Affinity designer. I currently use Pixelmator, which I am generally happy with. However I've recently purchased some fonts that require Opentype to use the ligatures and alternates. Before I purchase Affinity designer, will I be able to access the extras for the fonts purchased in this programme? I've searched everywhere in Pixelmator, and I don't think it supports it.

Really appreciate your advice :)

best wishes

Thea

[MikeW]

Hello Thea,

 

I've never used Pixelmator so I have no idea about it in any manner.

 

As for AD, a cautious Yes as regards it being able to utilize your OT font to its fullest. I wrote "a cautious Yes" because I am pretty certain there are fonts out in the wild that may cause issues in any application or particular ones.

 

I have not personally had issues other than a certain OT Feature being fully able to process--which is why this thread was put off for a month or so. The issue was the full processing of the rules I wrote for the medial (long) s character. AD wouldn't fully process one part of the complicated rule. So for instance, in the screen shots above there is the word, "crossstitch." the rule for using the medial s is that in an instance where there are 3 consecutive occurances, the middle one would be the normal s we all use. AD didn't process that rule completely.

 

So I wrote to the Serif developer who deals with this, he found the issue and it was rectified it in the very next beta and so in the first Windows release (which also applies to the Mac version you use). Which is the point I am wanting to get across: Serif is keen to rectify any issues.

 

And note there are different ways to write OT Feature code. Some fonts (including some of my own) utilize features in a manner that has now been redefined by the committee that deals with such things. These features (that control the initial word start characters, the middle ones and the endings of words) will be on by default. But they can be shut off in AD. It's just annoying initially is all.

 

Anyway, yes, your OT font should work fine.