Border Ornaments and their implementation: questions & answers.

Matthijs HerzbergMatthijs Herzberg Posts: 143
edited October 2020 in Technique and Theory
I'm working on a typeface to which I'd like to add ornamental border elements, as well as some other decorative glyphs.
In researching the best way to do so, I came across this thread, which lays out multiple options, but doesn't come to a conclusion on which method is preferable.

As I understand it, there's three options:
  1. A separate font file, with ornaments in place of standard keyboard characters (abc, 123, etc). Advantage: easy to use, compatible with even the most basic software. Disadvantage: wouldn't hold up well across applications, less compact (especially an issue in a variable font).
  2. Give ornaments their own UniCode (for instance in place of dingbat block). Map ornaments to keyboard characters, and have user activate them through an OpenType feature: there's apparently a dedicated Ornm feature, altough I've never come across it in any font. Alternatively, they can be added in a stylistic set.  Advantage: easy to use for people who know what they're doing. Disadvantage: relies on software supporting OpenType, can be annoying if working in a text block and unintentionally changing parts of the text to ornament.
  3. Give ornaments their own UniCode, and let users manually select them through a glyph panel/character insert tool. Advantage: should work in most software, each character retains appropriate unicode, best chance to work cross-application. Disadvantage: annoying to use.
None of these seem clearly better than the others, so I asked myself instead: what do users want?
Fortunately, I have a small social media following consisting largely of designers and design enthusiasts, so I asked them what they like using best through a poll. While hardly a scientific study, nearly 50 people answered.

I asked two questions, and got the following responses:
1. "Graphic Designers/Design Enthusiasts who use or have used border ornaments in fonts, do you prefer to have a separate font file with the ornaments in place of letters, or have them as unique characters within a single Font File?"
20 People voted for "Separate File" (42%)
28 People voted for "Same File" (58%)
2. "If within the same file, do you prefer to have it implemented through an OpenType feature, or do you manually select through Glyphs Panel?"
16 People voted for OpenType (35%)
30 People voted for Manual (65%)
Of course this is a very small sample size, but what did become clear is that there is no decisive opinion among designers on what the best implementation of border ornaments should be.
The good news: they're not mutually exclusive! I will simply use all three methods, and let users decide themselves how to handle it.
I'm sharing this with you all hoping that it may be of some use, and welcome your thoughts and personal experiences.

Comments

  • One more note: @Ray Larabie mentioned in the thread linked above, that he implemented dingbats/ornaments by having users type, for instance, ((lighting)), and OpenType would automatically sub it for the corresponding glyph. This seems like a smart way to do it, especially for less designer-y types, but does still depend on an application supporting OpenType. It would also be less useful when you have 50 abstract border elements, compared to easily describable dingbats.
  • Hi Matthijs, I have a thing or two to say about borders in particular. Twenty years ago I made very simple borders that consisted of eight glyphs: four corners, and one each for left, right, top, bottom. These mapped neatly to the keys on your keypad, so that you could easily type out a border. 7,8,9 formed the top row, 4 the left side, 5 was a matching space, and so forth. I made a separate font for each border. This way, you could produce a border in Notepad if you had to, and it was very easy to explain. My target audience has changed since then, and my borders have grown more sophisticated, so I, too, have cast about for better ways of doing this. 

    Here's what I've found: 

    1) Anything you squirrel away in stylistic sets sees very little use. People generally don't pay attention to the pretty PDF specimen you may have included, and unless they're font nerds, they rarely snoop around for opentype goodies. If it shows up in their font menu, they're more likely to use it. Moreover, if they're using Word or Powerpoint, they ain't never gonna see those features. 

    2) Separate fonts work well. I've gone so far as to make a single family, and then have a bunch of borders be different styles in that family. That works wonderfully, but only on Mac. If it ain't regular, bold or italic, Windows doesn't want to know about it. (I would love to be mistaken about this - anyone know how to get around that?) Also, especially with complex/multicolor borders. I wouldn't want to put several borders that have fifty or more elements each into a single font; it would make the glyph browser unusable. 

    Incidentally, a lot of folks poo-poo the idea of border fonts on principle. "Just make it a brush, bro". I would, only not everybody uses Illustrator. 

    Until something better comes along, I'll continue putting borders (and ornaments) into separate fonts. 

    Cheers,

    Oliver
  • edited October 2020
    Oh, and since I'm working on Art Nouveau borders this very minute, I figure I'd share this one, though you didn't ask for it: 





  • Matthijs HerzbergMatthijs Herzberg Posts: 143
    edited October 2020
    @Oliver Weiss (Walden Font Co.) Thanks for your hard-earned insight. I am inclined to agree with you on ornaments being more likely to be used if presented in a separate file, but I did have 28/48 people tell me they prefer to have them in the same file. Perhaps it depends on the target audience.

    Also wonderful job on that Art Nouveau border, I’m sure that took a lot of elbow grease.
  • Thank you, Matthijs 😄

    I was referring to complex borders specifically. Good organization is certainly a must, but even so, they do get hard to pick out when you have a great number of similar glyphs. 

    Ornaments/cuts/gems tend to be more distinctive, so not nearly as big a concern. And if they were designed for a particular typeface, it certainly makes sense to stick them in the same font. 

    • Give ornaments their own UniCode (for instance in place of dingbat block). 

    If you assign unicode values to your borders, you would want to use codepoints from one of the private use areas. Codepoints in the dingbats block (or any other block) already have expected glyph shapes associated with them.

  • If you assign unicode values to your borders, you would want to use codepoints from one of the private use areas. Codepoints in the dingbats block (or any other block) already have expected glyph shapes associated with them.
    Private Use Area codepoints also bear potential expectations about their content. I don’t see any harm by hijacking some of the “Dingbats” 2700 block, at least for similar elements. This block is nonsense anyway, so I feel rather free to assign a sensible task to it, if any.

Sign In or Register to comment.